BRCA2 c.3270G>C, p.(Met1090Ile)

Reasoning

Governing rule (ENIGMA BRCA1/BRCA2 specification v1.2, Table 1 / Appendix G): PM2_Supporting is 'Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer)'; the region around the variant must have an average read depth >=25, the variant must not be an insertion/deletion/delins (low recall), and observation of a variant only once in a gnomAD outbred population is not informative. This variant is a single-nucleotide substitution (NM_000059.4:c.3270G>C) and is reported ABSENT from both governing control sources - gnomAD v2.1 exomes (and its non-cancer subset) and gnomAD v3.1 non-cancer genomes. It is not a well-established pathogenic founder variant. Appendix G estimated the absent-in-controls LR towards pathogenicity at 2.88 (95% CI 2.49-3.34), which calibrates to Supporting weight. Context: gnomAD v4.1 all-comers reports very low-level presence (2/1,588,408 alleles, AF 1.25912e-06, 0 homozygotes; exome 2/1,436,146, genome 0/0), which does not contradict absence in the spec-mandated sources and remains well below any benign-frequency threshold. PM2 is met at Supporting strength.

Rule

PM2 (Supporting) - Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer). Region around the variant must have an average read depth >=25. Not applied to insertion/deletion/delins variants. See Appendix G.

Evidence considered

• ENIGMA specification v1.2, PM2 (default assigned weight: Supporting): 'Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer). Region around the variant must have an average read depth >=25. See Appendix G for details.'
• ENIGMA specification v1.2 instructions for PM2: 'Observation of a variant only once in a gnomAD outbred population is not informative. Do not apply for insertion, deletion or delins variants. Do not apply if read depth <25 at region around the variant.' This variant is a single-nucleotide substitution, so the indel exclusion does not apply.
• Appendices v1.2, Appendix G: absent-in-controls calibrated to PM2_Supporting (175/609 likely-benign vs 116/140 likely-pathogenic variants, LR 2.88, 95% CI 2.49-3.34); read depth >25 recommended for PM2_Supporting.
• gnomAD v2.1 (13-32911762-G-C): absent; gnomAD v2.1 non-cancer subset: absent; gnomAD v3.1 non-cancer: absent.
• gnomAD v4.1 all-comers (13-32337625-G-C): total AF 1.25912e-06 (2/1,588,408 alleles), homozygotes 0; exome 2/1,436,146 (AF 1.39262e-06); genome 0/0; best subpopulation Remaining individuals AF 3.26435e-05 (2/61,268). Recorded as context; it does not alter the governing-source absence.

Still missing

• Regional read depth >=25 (required by the ENIGMA v1.2 PM2_Supporting rule) was not explicitly captured in the case evidence. Absence across gnomAD v2.1 exomes and the presence of 2 called alleles among 1,436,146 v4.1 exome alleles at this site indicate the locus is mappable/well covered, but the explicit depth QC is not documented.
• gnomAD-Canada v1.0 lookup errored/timed out and returned no data.
• No VCEP-file entry for this exact variant was found (Specifications_V1.2_2024-11-18.docx, Appendices_V1.2_2024-11-18.docx, HUMU-40-1557-s001.xlsx searched for c.3270G>C / p.Met1090Ile / p.M1090I; no match), so no pre-assigned PM2 code exists to override this assessment.

Sources

Reasoning

The ENIGMA BRCA2 specification applies BP1_Strong for a silent, missense, or in-frame insertion/deletion variant located outside a (potentially) clinically important functional domain, provided no splicing is predicted (SpliceAI <= 0.1); missense prediction is explicitly not applicable to this assessment. NM_000059.4:c.3270G>C is a missense change, p.(Met1090Ile), and residue 1090 lies outside both BRCA2 (potentially) clinically important functional domains defined by ENIGMA (PALB2 binding domain aa 10-40 and DNA binding aa 2481-3186). Residue 1090 also lies outside the additional conserved region aa 1103-1183 noted in Appendix C, and the Appendix C catalogue of clinically important BRCA2 residues lists no residue at or near 1090. The case SpliceAI maximum delta is 0.0, comfortably at or below the 0.1 threshold, so no splicing is predicted. Both conditions for BP1_Strong are therefore satisfied. Independently consistent region-level support comes from the published BRCA2 coldspot analysis of exons 10 and 11 (codons 266-2281, which contains codon 1090), which found no pathogenic or likely pathogenic missense variants in that region (0/34) against 62.1% of benign/likely benign variation (odds ratio <0.01, 95% CI 0.0-0.01), satisfying criteria for strong benign evidence. BP1 is met at strong strength.

Rule

Apply BP1_Strong for a missense variant outside a (potentially) clinically important BRCA2 functional domain (PALB2 binding aa 10-40; DNA binding aa 2481-3186) when no splicing is predicted (SpliceAI <= 0.1); residue 1090 is outside the domains and SpliceAI max delta is 0.0.

Evidence considered

• ENIGMA Table 1 (Specifications v1.2) and Appendix J Table 16 direct that BP1_Strong be applied for silent, missense or in-frame insertion/deletion variants outside a (potentially) clinically important functional domain with no splicing predicted (SpliceAI <= 0.1).
• ENIGMA Appendix C (Table 4) and Appendix J define the (potentially) clinically important BRCA2 functional domains as PALB2 binding aa 10-40 and DNA binding aa 2481-3186; residue 1090 (Met1090Ile) is outside both, and Appendix C notes the additional conserved region aa 1103-1183 with no pathogenic missense substitutions recorded.
• The case evidence reports SpliceAI maximum delta 0.00 for c.3270G>C, satisfying the SpliceAI <= 0.1 no-splicing condition of BP1_Strong.
• PMID:31911673 defines BRCA2 exons 10 and 11 (codons 266-2281, encompassing codon 1090) as a coldspot: 0/34 pathogenic or likely pathogenic missense variants with an odds ratio for pathogenicity of <0.01 (95% CI 0.0-0.01), described by the authors as satisfying criteria for strong benign evidence.

Still missing

• No exact c.3270G>C, p.Met1090Ile, or p.M1090I entry found in Specifications_V1.2_2024-11-18.docx.txt.
• No exact c.3270G>C, p.Met1090Ile, or p.M1090I entry found in Appendices_V1.2_2024-11-18.docx.txt.
• No exact c.3270G>C, p.Met1090Ile, or p.M1090I entry found in Specifications_Table4_V1.2_2024-11-18.xlsx.txt; the governing BP1_Strong rule is in the specification text and Appendix J, not in that exon-level table.

Sources

Reasoning

The ENIGMA BRCA2 specification applies PS1 only when the variant is predicted to lead to the same missense substitution as an established pathogenic (or, at reduced weight, likely pathogenic) variant arising from a different nucleotide change, with no predicted or confirmed effect on mRNA splicing for either variant. NM_000059.4:c.3270G>C is predicted to yield NP_000050.3:p.(Met1090Ile), and its SpliceAI maximum delta is 0.0, so the splicing stipulation is satisfied, but no previously classified pathogenic or likely pathogenic variant producing p.Met1090Ile via an alternate nucleotide change could be identified. The only variant at this residue with any recorded classification is this exact change itself (ClinVar variation 51441), which carries conflicting interpretations (three laboratories Uncertain significance, two Likely benign; one-star review status) and therefore cannot serve as a pathogenic reference. No exact c.3270G>C, p.Met1090Ile, or p.M1090I entry exists in any of the three declared ENIGMA VCEP files. PS1 is therefore not met.

Rule

PS1 applies only when the variant leads to the same missense substitution as an established pathogenic (or likely pathogenic) variant due to a different nucleotide change, with no predicted/confirmed splicing effect; no such pathogenic reference exists here.

Evidence considered

• ENIGMA Appendix J states that PS1 may be applied for a variant predicted to lead to the same missense substitution as a known pathogenic variant arising from a different nucleotide change, at moderate weight if the reference variant is likely pathogenic, with no predicted or confirmed mRNA splicing effect.
• The normalized consequence is NM_000059.4:c.3270G>C, NP_000050.3:p.(Met1090Ile); the reported SpliceAI maximum delta is 0.0, satisfying the no-splicing stipulation.
• ClinVar variation 51441 for NM_000059.4(BRCA2):c.3270G>C (p.Met1090Ile) has conflicting classifications (3 Uncertain significance, 2 Likely benign) with one-star review status; no pathogenic or likely pathogenic variant producing p.Met1090Ile from a different nucleotide change was found.

Still missing

• No exact c.3270G>C, p.Met1090Ile, or p.M1090I entry found in Specifications_V1.2_2024-11-18.docx.txt.
• No exact c.3270G>C, p.Met1090Ile, or p.M1090I entry found in Appendices_V1.2_2024-11-18.docx.txt.
• No exact c.3270G>C, p.Met1090Ile, or p.M1090I entry found in Specifications_Table4_V1.2_2024-11-18.xlsx.txt.
• No alternative nucleotide change producing p.Met1090Ile with an established pathogenic or likely pathogenic classification was identified in ClinVar.

Sources

Reasoning

PS3 requires well-established in vitro or in vivo functional studies supportive of a damaging effect on protein function (ENIGMA BRCA1/BRCA2 v1.2 Specifications, Table 1; pre-assigned code weights in Specifications Table 9). NM_000059.4:c.3270G>C is a missense substitution, p.(Met1090Ile), and SpliceAI max delta 0.00 indicates no splice-mediated mechanism, so only protein-level (or combined mRNA+protein) assay evidence would qualify. The governing ENIGMA v1.2 curated functional-assay table (Specifications_Table9_V1.2_2024-11-18.xlsx), which lists the variants for which calibrated assays have been reviewed and a PS3/BS3 code weight assigned, contains no row for this variant under any notation searched (c.3270G>C, p.Met1090Ile, p.M1090I, M1090I, or codon 1090). No functional-assay result for BRCA2 p.Met1090Ile was found in the Parsons et al. 2019 multifactorial dataset (HUMU-40-1557-s001.xlsx, which cross-references functional-assay data) or in SupplementaryTables_V1.2_2024-11-18.xlsx. OncoKB reports no variant-specific reviewed functional evidence for this variant. None of the four full-text publications fetched for this case (PMIDs 12692171, 23918944, 31672839, 31911673) names the variant or reports an assay on it. No damaging-effect functional evidence therefore exists and PS3 is not met.

Evidence considered

• Governing rule consulted: ENIGMA BRCA1/BRCA2 v1.2 PS3 - 'Well-established in vitro or in vivo functional studies supportive of a damaging effect. Apply PS3 for assays measuring effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E.' Table 9 states that additional published assay results not listed there should be reviewed following Figure 1C before applying PS3/BS3.
• Searched the full BRCA2 VCEP full-text corpus (Specifications_V1.2, Appendices_V1.2, Specifications_Table4, Specifications_Table9, SupplementaryTables, HUMU-40-1557-s001) with grep for 'c.3270G>C', 'p.Met1090Ile', 'p.M1090I', 'M1090I', 'Met1090' and '1090'. No BRCA2 record matched; the only 1090 matches were BRCA1 entries (e.g. BRCA1 c.3270A>T p.Gln1090His), a different gene and variant.
• Specifications_Table9 (4765 lines, ~533 BRCA2-containing rows) is the on-record table of curated functional assays with pre-assigned PS3/BS3 weights; it covers BRCA2 variants across the gene including many missense changes, but codon 1090 / c.3270 is absent.
• Variant consequence context: missense (NP_000050.3:p.(Met1090Ile)); SpliceAI max delta 0.00, so PS3/BS3 here can only rest on protein-level (or combined mRNA+protein) assay evidence, not on a splice effect.
• OncoKB (BRCA2 M1090I) reported no variant-specific reviewed functional evidence; only gene-level BRCA2 DNA-damage-response context was available.
• Dines et al. 2020 (PMID:31911673) places BRCA2 codon 1090 inside the exons 10-11 'coldspot' (codons 266-2281) reported as tolerant of missense variation, but that work is a Bayesian reanalysis of ClinVar classifications, not an in vitro/in vivo functional assay, so it is not PS3 evidence and was not counted.

Still missing

• Specifications_Table9_V1.2_2024-11-18.xlsx (the ENIGMA BRCA1/BRCA2 v1.2 curated functional-assay table that carries pre-assigned PS3/BS3 weights) was searched for 'c.3270G>C', 'p.Met1090Ile', 'p.M1090I', 'M1090I', '1090' and the hg38 position: no entry exists for this variant. A missing Table 9 entry is not by itself proof that no functional assay has ever been published, since the table only catalogues assays the VCEP has reviewed and calibrated.
• HUMU-40-1557-s001.xlsx (Parsons et al. 2019, PMID:31131967; multifactorial dataset with a functional-assay cross-reference) and SupplementaryTables_V1.2_2024-11-18.xlsx were searched: neither contains a functional-assay result for BRCA2 codon 1090.
• No BRCA2 p.Met1090Ile functional assay (e.g. homology-directed repair, saturation genome editing, mouse ES-cell viability, or combined mRNA+protein assay) was identified in the four full-text publications retrieved for this case; each was screened for the variant and none mentions it. Assays not captured by this case's retrieval (un-indexed or not variant-titled papers) remain a residual gap.

Sources

Reasoning

PS4 requires that the prevalence of the variant in affected individuals be significantly increased compared with controls. For BRCA2 the governing framework is the ENIGMA v1.2 specification: 'The prevalence of the variant in affected individuals is significantly increased compared with the prevalence in controls. Case-control studies; p-value <=0.05 and OR >=4 (lower confidence interval excludes 2.0).', and Appendix F restates this as a relative risk or odds ratio >=4.0 whose confidence interval does not include 2.0, with ethnicity- and country-matched controls; proband counting ('PS4_Moderate') is explicitly 'Do not use'. Every framework source declared for PS4 was searched for this variant in all three notations (c.3270G>C, p.Met1090Ile, p.M1090I): the SupplementaryTables v1.2 ST7 Reference Set Vts sheet (reference-set variants with prior/posterior probability, IARC class and gnomAD frequencies), the Parsons et al. 2019 multifactorial tables (HUMU-40-1557-s001, SuppT1 and SuppT5), Specifications Table 11 ('Variants assessed as potential moderate-risk from case-control analysis'), and Appendix F. None of these contains an entry for this variant. The only hit for the search string 'c.3270' in these files is the unrelated BRCA1 variant c.3270A>T p.(Gln1090His). Consequently no variant-level odds ratio, relative risk, or case-control likelihood ratio exists for BRCA2 c.3270G>C, and no fetched publication reports case-control counts for it. The ClinVar record (VCV000051441) contains only single-laboratory assertions (3 laboratories Uncertain significance, 2 laboratories Likely benign), not case-control data. Population data reinforce that no single-variant enrichment estimate is possible: the variant is absent from gnomAD v2.1 and totals 2/1,588,408 alleles in gnomAD v4.1 (AF 1.259e-06, exome only, 0 homozygotes). Operator check as required before concluding: the governing operator for PS4 is '>=' (OR >=4.0); no OR value exists for this variant, so the inequality 'OR >= 4.0' cannot be instantiated with a substituted value, and with no excess of the variant in cases PS4 is not met.

Rule

ENIGMA BRCA2 v1.2 PS4: prevalence of the variant in affected individuals significantly increased vs controls; case-control studies p<=0.05 and OR >=4.0 with lower confidence interval excluding 2.0 (Appendix F); PS4_Moderate proband counting not to be used.

Evidence considered

• ENIGMA BRCA2 v1.2 Table 1 / Appendix F define PS4 as case-control enrichment with p<=0.05 and OR >=4 (lower CI excludes 2.0); Appendix F adds that case-control LR estimates, when available for a dataset, are to be applied under PP4 or BP5 in preference to a case-control OR/RR, and that proband counting ('PS4_Moderate') is not to be used.
• Searched every PS4-declared framework source for the variant under all three notations (c.3270G>C, p.Met1090Ile, p.M1090I): no entry exists. The only 'c.3270' match in these files is BRCA1 c.3270A>T p.(Gln1090His), a different gene and different variant.
• No odds ratio, relative risk, case-control LR or case-control counts for this variant appear in any fetched publication; the ClinVar submissions are single-laboratory classifications without case-control data.
• gnomAD v4.1: total AF 1.2591223413631762e-06 (2/1,588,408 alleles, exome only, 0 homozygotes), highest subpopulation AF 3.264e-05 in 'Remaining individuals'; absent from gnomAD v2.1.
• Operator direction check for this criterion specifically: the PS4 operator as literally written in the governing ENIGMA text is '>=' (OR >=4.0), i.e. met requires the value to go UP past the threshold; with no value available the inequality cannot be written with a substituted number, so PS4 is not met.

Still missing

• SupplementaryTables_V1.2.xlsx is declared by the ENIGMA BRCA2 index for PS4 but its full text is unavailable (could not be converted); the same-content conversion SupplementaryTables_V1.2_2024-11-18.xlsx.txt was searched instead and contains no entry for c.3270G>C / p.(Met1090Ile) / p.M1090I.
• No entry for c.3270G>C (p.Met1090Ile) in SupplementaryTables_V1.2_2024-11-18.xlsx 'ST7 Reference Set Vts' (searched, file available and read).
• No entry for c.3270G>C in HUMU-40-1557-s001.xlsx (Parsons et al. 2019, PMID:31131967) SuppT1/SuppT5 (searched, file available and read).
• No variant-level case-control dataset reporting BRCA2 c.3270G>C was found in any fetched publication; PMID:35264596 (cited by the Ambry Genetics ClinVar submission) has no retrievable full text in this case, only a screenshot.

Sources

Reasoning

ENIGMA BRCA2 specification v1.2 restricts PM3_VariableWeight to probands with a phenotype consistent with BRCA2-related (FANCD1) Fanconi Anemia who carry a co-occurrent pathogenic/likely pathogenic variant in BRCA2. Points are assigned per proband (2 points for a P/LP variant in trans or a homozygote, 1 point when phase is unknown) and the code is only assigned when the summed points across unrelated probands reach >=2 (PM3=2 points; PM3_Strong>=4; PM3_Supporting=1). This variant c.3270G>C (p.Met1090Ile) is a heterozygous missense observed in breast/ovarian cancer clinical testing: the ClinVar record (variation 51441) contains only heterozygous HBOC submissions (3 laboratories Uncertain significance, 2 laboratories Likely benign) with no Fanconi Anemia phenotype, no co-occurring BRCA2 P/LP variant, and no homozygote. No trans/cis phase data, no second-variant observation, and no FA-affected proband were found in the ClinVar submission audit or in any retrieved publication (all four full-text papers checked explicitly searched for c.3270G>C/p.M1090Ile and none reported the variant or any allelic co-occurrence). With zero PM3-relevant proband observations the summed points are 0, below even the PM3_Supporting threshold of 1. PM3 is not met.

Rule

PM3_VariableWeight (ENIGMA BRCA2 v1.2): applied only for a BRCA2-related Fanconi Anemia proband with a co-occurrent BRCA2 P/LP variant; PM3_Strong >=4 points, PM3 = 2 points, PM3_Supporting = 1 point.

Evidence considered

• ENIGMA BRCA1/BRCA2 specification v1.2 (Specifications_V1.2_2024-11-18.docx) states PM3_VariableWeight applies only to a patient with a phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia and co-occurrent variants in the same gene, and that PM3_Strong=4 points, PM3=2 points, PM3_Supporting=1 point (Table 6: 2 points for a P or LP variant in trans or a homozygote, 1 point when phase is unknown, per FA-affected proband).
• CSPEC criteria list for BRCA2 (version 1.2) marks PM3 as applicable with default strength Pathogenic Moderate; BP2 is marked 'Not Applicable'.
• ClinVar variation 51441 (c.3270G>C, p.Met1090Ile) has 6 germline submissions: 3 laboratories report Uncertain significance and 2 report Likely benign (plus one BIC record without criteria). All are clinical HBOC testing in heterozygotes; no Fanconi Anemia phenotype, no co-occurring BRCA2 pathogenic variant, and no homozygote are recorded.
• ClinVar submission audit for this exact variant lists no criteria_found for PM3/BP2 and no phase (trans/cis) information in any submission.
• Literature checked for allelic/co-occurrence evidence: PMID:23918944 (tamoxifen/CBC cohort), PMID:31672839 (pancreatic surveillance guideline), PMID:31911673 (BRCA2 coldspot reanalysis) and PMID:12692171 (ASCO policy statement) all explicitly searched for c.3270G>C/p.M1090Ile and none names the variant or reports any second BRCA2 variant, phase, or FA patient.

Still missing

• No BRCA2-related Fanconi Anemia proband carrying this variant, and no co-occurring pathogenic/likely pathogenic BRCA2 variant in trans/cis, was identified in the ClinVar submission audit or in the retrieved literature; PM3 requires such an observation (>=2 points) and none exists.
• Searched the full BRCA2 VCEP full-text files Specifications_V1.2_2024-11-18.docx, Appendices_V1.2_2024-11-18.docx, SupplementaryTables_V1.2_2024-11-18.xlsx, Specifications_Table4_V1.2_2024-11-18.xlsx, Specifications_Table9_V1.2_2024-11-18.xlsx, HUMU-40-1557-s001.xlsx and PMID_31853058_BRCA2_clinical_history_LR.xlsx for c.3270G>C / p.Met1090Ile / p.M1090I; there is no entry for this variant (no pre-assigned PM3 code), so the score was derived from the governing rule rather than a table lookup.

Sources

Reasoning

NM_000059.4:c.3270G>C is a missense substitution (NP_000050.3:p.(Met1090Ile)), so PP3 is evaluated on the protein-impact path, which for BRCA2 is governed by the ENIGMA BRCA1/BRCA2 specification v1.2 (cspec) and uses BayesDel (no-AF) rather than REVEL. The BRCA2 PP3 missense rule requires BOTH (i) location inside a (potentially) clinically important functional domain and (ii) a predicted impact via protein change, BayesDel no-AF >= 0.30 (VCEP specification Table 1 and Appendix J Table 16). Residue 1090 falls outside both BRCA2 clinically important functional domains, defined as the PALB2-binding domain (aa 10-40) and the DNA-binding domain (aa 2481-3186), so condition (i) fails. Condition (ii) also fails: the local BayesDel no-AF score for this variant is -0.499525, far below the >= 0.30 impact threshold (and below the <= 0.18 no-impact cutpoint, i.e. it sits in the indeterminate zone on the calibrated BRCA2 scale). Appendix J explicitly states that for missense variants outside a clinically important functional domain the missense prediction is not applicable and the small benign prior is instead captured by BP1 (BP1_Strong), not PP3. The VCEP additionally permits a splice-based PP3 route (SpliceAI >= 0.2) for missense variants irrespective of domain location; SpliceAI max delta for this variant is 0.00, below that threshold, so no splice-based PP3 either. A generic REVEL score of 0.319 is likewise in the indeterminate zone (PP3-supporting requires >= 0.644, Pejaver et al. 2022, PMID:36413997) and is not the governing tool for BRCA2 missense. No computational path reaches the PP3 threshold; PP3 is not met.

Rule

ENIGMA BRCA2 PP3 (v1.2): missense/in-frame inside a clinically important functional domain with BayesDel no-AF >= 0.30; OR predicted splicing with SpliceAI >= 0.2 (silent, missense/in-frame irrespective of domain, or intronic outside donor/acceptor +/-1,2). Note: do not apply PP3 if PVS1 is met at any strength.

Evidence considered

• Variant consequence: missense substitution NM_000059.4:c.3270G>C, NP_000050.3:p.(Met1090Ile) - therefore the protein-impact (missense) path, not the splice path, supplies PP3/BP4.
• ENIGMA BRCA1/BRCA2 specification v1.2, Table 1 (PP3): 'Apply PP3 for missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain and predicted impact via protein change (BayesDel no-AF score >= 0.30)'; 'Apply PP3 for predicted splicing (SpliceAI >= 0.2) for silent, missense/in-frame (irrespective of location in clinically important functional domain) and for intronic variants outside of donor and acceptor 1,2 sites.'
• ENIGMA v1.2 footnote: '(potentially) clinically important functional domains ... BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186.' Residue Met1090 lies outside both intervals.
• Appendix J, Table 16 (PP3 row): 'Apply PP3 for missense or in-frame insertion-deletion inside a (potentially) clinically important functional domain and predicted impact via protein change (BayesDel >= 0.28 for BRCA1, >= 0.30 for BRCA2).'
• Appendix J, Table 16 (BP1 row and PM1 row): missense outside a clinically important functional domain is evaluated under BP1/BP4/PP3 with 'Missense prediction not applicable' - the out-of-domain missense prior is captured by BP1_Strong, not by PP3.
• BayesDel (no-AF) local lookup for this variant: -0.499525, below the BRCA2 PP3 impact cutpoint (>= 0.30) and below the BRCA2 no-impact cutpoint (<= 0.18).
• SpliceAI local lookup for NM_000059.4:c.3270G>C: DS_AG 0.00, DS_AL 0.00, DS_DG 0.00, DS_DL 0.00; max delta 0.00, below the VCEP splice-based PP3 threshold of >= 0.2.
• REVEL local lookup for this variant: 0.319 - indeterminate zone; generic PP3-supporting requires >= 0.644 (Pejaver et al. 2022, Am J Hum Genet, PMID:36413997). REVEL is not the ENIGMA BRCA2-governing missense tool.
• Dines et al. 2020 (PMID:31911673) describes BRCA2 exons 10-11 (codons 266-2281) as a missense 'coldspot' - codon 1090 lies in that interval - but the paper does not name this variant and its coldspot proposal is not the ENIGMA-calibrated rule for PP3.

Still missing

• Searched the governing VCEP files listed in the index for this exact variant and found no variant-specific pre-assigned PP3 entry: Specifications_V1.2_2024-11-18.docx, Appendices_V1.2_2024-11-18.docx, PMID_22505045_Houdayer_2012_HumMutat.pdf, SupplementaryTables_V1.2_2024-11-18.xlsx, Specifications_Table4_V1.2_2024-11-18.xlsx, Specifications_Table9_V1.2_2024-11-18.xlsx, PMID_31853058_BRCA2_clinical_history_LR.xlsx (searched c.3270G>C, p.Met1090Ile, p.M1090I, M1090). No entry exists, so PP3 is decided by the VCEP rule rather than a pre-assigned code.

Sources

Reasoning

PP4 captures combined clinical evidence towards pathogenicity from multifactorial likelihood data. The ENIGMA BRCA2 v1.2 specification states 'Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data' and sets PP4 at LR >=2.08:1 (supporting), LR >=4.3:1 (moderate), LR >=18.7:1 (strong) and LR >=350:1 (very strong), with a combined LR of 1.00 to <2.08 explicitly non-informative ('PP4 not applicable'). The variant-level governing sources were searched for c.3270G>C, p.Met1090Ile and p.M1090I: the Li et al. 2020 (PMID:31853058) BRCA2 clinical-history likelihood-ratio table (the gene-specific PP4/BP5 table, columns GENE / VARIANT / HGVS_Nucleotide / LOG(LR) / N_Probands / LR = exp(LOG(LR))) contains no entry for this variant; the Parsons et al. 2019 multifactorial table (HUMU-40-1557-s001 SuppT1, columns Segregation LR / Pathology LR / Co-occurrence LR / Family History LR / Case-Control LR / Combined LR) contains no entry for this variant; and the Easton et al. 2007 combined multifactorial assessment (PMID:17924331, which names odds in favour of causality for individual VUSs) does not list it either. There is therefore no combined clinical LR for this variant in any governing or fetched source. Operator check as required before concluding: the governing operator for PP4 is '>=' (supporting LR >=2.08); no LR value exists for this variant, so 'LR >= 2.08' cannot be instantiated with a substituted number, and PP4 is not met.

Rule

ENIGMA BRCA2 v1.2 PP4: combined clinical multifactorial LR towards pathogenicity - supporting LR >=2.08:1, moderate >=4.3:1, strong >=18.7:1, very strong >=350:1; combined LR 1.00 to <2.08 not informative.

Evidence considered

• ENIGMA BRCA2 v1.2 PP4 rule text: 'Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data. PP4_Strong - LR >=18.7:1; PP4_Very Strong - LR >=350:1'; PP4_Moderate LR >=4.3:1; PP4 (supporting) LR >=2.08:1; 'Combined LR 1.00 to <2.08 is not informative (PP4 not applicable)'.
• The Li et al. 2020 BRCA2 clinical-history LR table was searched for c.3270G>C, p.Met1090Ile and p.M1090I and contains no row for this variant (rows are keyed by HGVS_Nucleotide, e.g. BRCA2 c.3437A>T, c.4133_4136delCTCA, c.4741G>A).
• The Parsons et al. 2019 SuppT1 multifactorial table was searched for the same notations and contains no row for this variant; the only '1090' matches in that file are unrelated BRCA1 variants and numeric values.
• Operator direction check for this criterion specifically: the PP4 operator as literally written in the governing ENIGMA text is '>=' (LR >=2.08 supporting), i.e. met requires the value to go UP past the threshold; no LR value is available to substitute, so the criterion is not met.

Still missing

• No entry for c.3270G>C / p.Met1090Ile / p.M1090I in PMID_31853058_BRCA2_clinical_history_LR.xlsx (the gene-specific PP4/BP5 clinical-history LR table; searched, file available and read).
• No entry for this variant in HUMU-40-1557-s001.xlsx SuppT1 (Parsons et al. 2019, PMID:31131967) combined multifactorial LR table; searched, file available and read.
• No entry for this variant in the PMID:17924331 (Easton et al. 2007) combined multifactorial assessment; the paper's variant-level tables do not name c.3270G>C/p.M1090I.
• No unpublished or published multifactorial clinical data (personal/family history, tumour pathology, co-occurrence, cosegregation) specific to this variant were identified anywhere in the case evidence.

Sources

Reasoning

PP5 is applied only from an exact-variant ClinVar expert-panel classification of Pathogenic or Likely pathogenic; ordinary laboratory submissions, aggregate ClinVar labels and non-expert assertions must never trigger it. ClinVar VCV000051441 (variation ID 51441, NM_000059.4(BRCA2):c.3270G>C p.Met1090Ile) contains six submissions from six different submitters: Ambry Genetics (Likely benign), Labcorp Genetics/Invitae (Likely benign), Breast Cancer Information Core (Uncertain significance, no assertion criteria provided), LabCorp Women's Health and Genetics (Uncertain significance), Color Diagnostics (Uncertain significance) and University of Washington Department of Laboratory Medicine (Likely benign). The submission audit records expert_panel_submissions = 0; the ClinVar review status is 'criteria provided, conflicting classifications' with 1 star; and the aggregate API label is 'Conflicting classifications of pathogenicity'. No ClinGen expert panel (ENIGMA BRCA1/BRCA2 VCEP) classification of this variant exists, so no submission qualifies as an exact-variant expert-panel Pathogenic/Likely pathogenic assertion. The ENIGMA BRCA2 v1.2 specification itself lists PP5 as 'Do not use'. PP5 is therefore not met.

Rule

Exact-variant ClinVar expert-panel (ClinGen VCEP) classification of Pathogenic or Likely pathogenic is required to apply PP5; ordinary laboratory submissions and aggregate ClinVar labels do not qualify. ENIGMA BRCA2 v1.2 lists PP5 as 'Do not use'.

Evidence considered

• ClinVar VCV000051441 / variation 51441: 6 submissions from 6 submitters, expert_panel_submissions = 0, review_stars = 1, review status 'criteria provided, conflicting classifications', aggregate label 'Conflicting classifications of pathogenicity' (3 laboratories Uncertain significance, 2 laboratories Likely benign).
• ENIGMA BRCA2 v1.2 specification lists PP5 as 'Do not use'; per the governing global rule PP5 is nonetheless evaluated for an exact-variant ClinVar expert-panel Pathogenic/Likely pathogenic classification, and none exists.

Still missing

• No ClinGen expert-panel (ENIGMA BRCA1/BRCA2 VCEP) ClinVar classification exists for VCV000051441; the record is 1-star with conflicting ordinary-laboratory submissions.

Sources

Reasoning

Governing rule (ENIGMA BRCA1/BRCA2 specification v1.2, Table 1 / Appendix G): BA1 (Stand Alone) requires filter allele frequency (FAF) > 0.001 (0.1%) in gnomAD v2.1 (non-cancer, exome-only subset) and/or gnomAD v3.1 (non-cancer), in non-founder population(s); the BA1 cutoff of 0.000906 (BRCA2) was derived with the Whiffin calculator (prevalence 0.125, genetic heterogeneity 0.01, allelic heterogeneity 1.00, penetrance 0.69) and rounded up to 0.001. This variant is reported ABSENT from gnomAD v2.1 exomes, from the gnomAD v2.1 non-cancer subset, and from gnomAD v3.1 non-cancer. In gnomAD v4.1 all-comers the variant is present at total AF 1.25912e-06 (2/1,588,408 alleles; exome 2/1,436,146, genome 0/0) with 0 homozygotes, and the highest observed subpopulation frequency is 3.26435e-05 (2/61,268 alleles, Remaining individuals) - still more than two orders of magnitude below the 0.001 stand-alone threshold. BA1 is therefore not met. The variant is a single-nucleotide substitution in a gene with no established pathogenic founder effect relevant here, so no founder exclusion applies.

Rule

BA1 (Stand Alone) - Filter allele frequency (FAF) is above 0.1% (FAF > 0.001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.

Evidence considered

• ENIGMA BRCA1/BRCA2 specification v1.2, BA1: 'Filter allele frequency (FAF) is above 0.1% (FAF > 0.001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.'
• Appendices v1.2, Appendix G: BA1 calculated with the Whiffin calculator (prevalence 0.125, genetic heterogeneity 0.01, allelic heterogeneity 1.00, penetrance BRCA2 0.69); minimal credible allele frequency 0.000906 for BRCA2 rounded up to 0.001. BA1/BS1 to be applied using filter allele frequency (not MAF), highest FAF in a gnomAD non-founder population, assessing exome and genome data separately, and not applied to well-established pathogenic founder variants.
• gnomAD v4.1 all-comers screen (13-32337625-G-C): total AF 1.25912e-06, 2/1,588,408 alleles, homozygotes 0; exome 2/1,436,146 (AF 1.39262e-06), genome 0/0; highest subpopulation Remaining individuals AF 3.26435e-05 (2/61,268 alleles, homozygotes 0).
• gnomAD v2.1 (13-32911762-G-C) and gnomAD v2.1 non-cancer subset: variant absent. gnomAD v3.1 non-cancer: variant absent.
• The gnomAD v4.1 dataset is not the source named by the governing v1.2 BA1 rule (which specifies gnomAD v2.1 non-cancer exome and/or gnomAD v3.1 non-cancer), so the all-comers figure is recorded as context only; even taken at face value it is far below the 0.001 threshold.

Still missing

• gnomAD-Canada v1.0 (HostSeq) lookup errored/timed out and returned no data; this source is not one of the ENIGMA v1.2 BA1 sources and its absence does not affect the verdict.
• No VCEP-file entry for this exact variant was found: Specifications_V1.2_2024-11-18.docx, Appendices_V1.2_2024-11-18.docx and HUMU-40-1557-s001.xlsx were searched for c.3270G>C / p.Met1090Ile / p.M1090I and contain no entry for it; the criterion was therefore evaluated directly against the specification rule text and gnomAD data.

Sources

Reasoning

Governing rule (ENIGMA BRCA1/BRCA2 specification v1.2, Table 1 / Appendix G): BS1_Strong requires FAF > 0.0001 (0.01%), and BS1_Supporting may be applied for FAF > 0.00002 (0.002%) and <= 0.0001, in gnomAD v2.1 (non-cancer, exome-only subset) and/or gnomAD v3.1 (non-cancer) non-founder population(s). The BRCA2 BS1 cutoff of 0.0000906 was derived with the Whiffin calculator (allelic heterogeneity 0.10, penetrance 0.69) and rounded up to 0.0001. This variant is ABSENT from gnomAD v2.1 exomes, gnomAD v2.1 non-cancer, and gnomAD v3.1 non-cancer, so neither BS1_Strong nor BS1_Supporting is met in the governing sources. Context: in gnomAD v4.1 all-comers the variant is seen twice (total AF 1.25912e-06, 2/1,588,408, 0 homozygotes); the highest subpopulation frequency (3.26435e-05, Remaining individuals 2/61,268) falls within the BS1_Supporting band (>0.00002 to <=0.0001), but (i) that observation is from the all-comers joint dataset, not the spec-mandated gnomAD v2.1 non-cancer exome or gnomAD v3.1 non-cancer sources, (ii) 'Remaining individuals' is not one of the non-founder populations enumerated by the specification (Non-Finnish European, African, Latino, East Asian, South Asian), and (iii) an allele count of 2 represents a single-occurrence class which Appendix G found uninformative. BS1 is therefore not met; human review is flagged because the newer all-comers data grazes the BS1_Supporting lower bound.

Rule

BS1 (Strong) - Filter allele frequency (FAF) is above 0.01% (FAF > 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). BS1 (Supporting) - FAF > 0.00002 and <= 0.0001 in the same sources. See Appendix G.

Evidence considered

• ENIGMA specification v1.2, BS1: 'Filter allele frequency (FAF) is above 0.01% (FAF > 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s)' (Strong); Supporting tier: 'FAF is above 0.002% (FAF > 0.00002) and less than or equal to 0.01% (FAF <= 0.0001)' in the same sources.
• Appendices v1.2, Appendix G: BS1 minimal credible allele frequency 0.0000906 (BRCA2) rounded up to 0.0001; LR-based analysis gave LR 0.03 (Benign Strong) for MAF >0.0001 to <=0.001 and supported a conservative BS1_Supporting for MAF >0.00002 to <=0.0001. BA1/BS1 to use filter allele frequency, highest FAF in a gnomAD non-founder population, exome and genome assessed separately, TCGA excluded.
• gnomAD v2.1 (13-32911762-G-C), gnomAD v2.1 non-cancer, and gnomAD v3.1 non-cancer: variant absent.
• gnomAD v4.1 all-comers (13-32337625-G-C): 2/1,588,408 alleles, AF 1.25912e-06, homozygotes 0; best subpopulation 'Remaining individuals' 2/61,268, AF 3.26435e-05 - recorded as context only because it is not the dataset named by the governing rule.
• Appendix G states that observation of a variant only once in a gnomAD outbred population is not informative (LR 0.98, 95% CI 0.59-1.63), and maximum MAF >0 to <=0.00002 was likewise not informative.

Still missing

• Borderline: gnomAD v4.1 all-comers shows 2 alleles with a maximum subpopulation AF of 3.26435e-05 ('Remaining individuals'), which exceeds the BS1_Supporting lower bound (>0.00002) but is outside the ENIGMA v1.2 enumerated non-founder population list and the spec-mandated non-cancer datasets; human review advised on whether newer all-comers data should update the v2.1/v3.1-based assessment.
• No VCEP-file entry for this exact variant was found (Specifications_V1.2_2024-11-18.docx, Appendices_V1.2_2024-11-18.docx, HUMU-40-1557-s001.xlsx searched for c.3270G>C / p.Met1090Ile / p.M1090I; no match), so no pre-assigned BS1 code exists to override this assessment.
• gnomAD-Canada v1.0 lookup errored/timed out (no data).

Sources

Reasoning

BS3 requires well-established in vitro or in vivo functional studies showing no damaging effect on protein function, with the assay measuring effect via protein only or via mRNA and protein combined (ENIGMA BRCA1/BRCA2 v1.2 Specifications, Table 1; pre-assigned code weights in Specifications Table 9). NM_000059.4:c.3270G>C (p.Met1090Ile) is a missense variant; no calibrated assay reporting a benign (wild-type-like) functional result exists for it. The governing ENIGMA v1.2 curated functional-assay table (Specifications_Table9_V1.2_2024-11-18.xlsx) has no row for c.3270G>C, p.Met1090Ile, p.M1090I or codon 1090, and no neutral-result assay for this variant was found in HUMU-40-1557-s001.xlsx, SupplementaryTables_V1.2_2024-11-18.xlsx, or the four full-text publications retrieved for this case. Regions described as 'coldspots' in the literature are not functional assays: Dines et al. 2020 (PMID:31911673) reports that BRCA2 exons 10-11 tolerate missense variation and proposes classifying such variants as likely benign under the benign-prediction criteria (BP1/BP4), but it performed no assay of protein function and therefore cannot be used as BS3 evidence. With no well-established non-damaging functional assay available, BS3 is not met.

Evidence considered

• Governing rule consulted: ENIGMA BRCA1/BRCA2 v1.2 BS3 - 'Well-established in vitro or in vivo functional studies shows no damaging effect on protein function. Assay measures effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E.'
• Searched the BRCA2 VCEP full-text corpus for this variant under all common notations; the only hit for '1090' was a BRCA1 record (BRCA1 c.3270A>T p.Gln1090His), not BRCA2 c.3270G>C.
• Specifications_Table9 rows carry an explicit 'Assigned Code' and 'Code Weight' column (e.g. 'BS3 / Strong' for several BRCA1 and BRCA2 missense variants); no such row exists for this variant, so there is no on-record curated neutral-result assay to apply.
• HUMU-40-1557-s001 (Parsons et al. 2019, PMID:31131967) cross-references functional-assay data but contains no BRCA2 p.Met1090Ile record; its BRCA2 rows near this region (e.g. c.3772A>G p.Ile1258Val) carry only multifactorial likelihood values, not calibrated assay outcomes.
• Dines et al. 2020 (PMID:31911673) was reviewed as a possible benign-direction source: it defines BRCA2 exons 10-11 (codons 266-2281, which include codon 1090) as a variation-tolerant 'coldspot' with an odds ratio for pathogenicity <0.01, but this is a ClinVar-based Bayesian analysis, not a functional assay, so it cannot support BS3 and was not counted as such.

Still missing

• Specifications_Table9_V1.2_2024-11-18.xlsx (ENIGMA BRCA1/BRCA2 v1.2 curated functional-assay table with pre-assigned PS3/BS3 weights) was searched for 'c.3270G>C', 'p.Met1090Ile', 'p.M1090I', 'M1090I', '1090' and the hg38 position: no entry exists for this variant, so no pre-assigned BS3 weight can be retrieved. Absence of a Table 9 row is a weaker statement than proof that no neutral-result assay has ever been published.
• No functional-assay result for BRCA2 p.Met1090Ile was found in HUMU-40-1557-s001.xlsx (PMID:31131967) or SupplementaryTables_V1.2_2024-11-18.xlsx; neither dataset contains a BRCA2 codon 1090 functional entry.
• The four full-text publications retrieved for this case (PMIDs 12692171, 23918944, 31672839, 31911673) were screened for the variant and none reports a functional assay of it; a benign functional result could therefore not be established from the available literature.
• A group of independent, calibrated mammalian assays (e.g. HDR, SGE, ES-cell viability) for BRCA2 p.Met1090Ile would be required to convert this to a met BS3; no such evidence was located.

Sources

Reasoning

c.3270G>C is a missense substitution (p.Met1090Ile), so BP4 is evaluated on the protein-impact path; for BRCA2 the governing ENIGMA v1.2 BP4 missense rule requires the variant to be BOTH inside a (potentially) clinically important functional domain AND to show no predicted impact via protein change or splicing (BayesDel no-AF <= 0.18 AND SpliceAI <= 0.1). Residue 1090 lies outside both BRCA2 clinically important functional domains (PALB2-binding aa 10-40; DNA-binding aa 2481-3186), so the domain condition of the BRCA2 missense BP4 rule is not satisfied and the rule cannot be met. Although the numeric no-impact conditions would be satisfied in isolation (BayesDel no-AF = -0.499525 <= 0.18; SpliceAI max delta 0.00 <= 0.1), Appendix J Table 16 states that outside a clinically important functional domain the missense prediction is not applicable and the benign bioinformatic prior for out-of-domain missense variants is applied through a different code (BP1_Strong, 'missense or in-frame insertion-deletion outside a (potentially) clinically important functional domain ... SpliceAI <= 0.1'), which is assessed by another group. Independently, the non-governing generic REVEL score of 0.319 falls in the indeterminate gray zone (BP4-supporting requires <= 0.29, Pejaver et al. 2022, PMID:36413997), so BP4 is not met on the generic path either. The clean SpliceAI result is not used as BP4 support for a missense variant. BP4 is not met.

Rule

ENIGMA BRCA2 BP4 (v1.2): missense/in-frame inside a clinically important functional domain with BayesDel no-AF <= 0.18 AND SpliceAI <= 0.1; or silent variant inside such a domain with SpliceAI <= 0.1; or intronic variant outside donor/acceptor +/-1,2 with SpliceAI <= 0.1.

Evidence considered

• Variant consequence: missense substitution NM_000059.4:c.3270G>C, NP_000050.3:p.(Met1090Ile).
• ENIGMA v1.2, Table 1 (BP4): 'Missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain, and no predicted impact via protein change or splicing (BayesDel predicted No Impact AND SpliceAI <= 0.1).' BRCA2 no-impact BayesDel cutpoint: <= 0.18.
• ENIGMA v1.2 footnote: clinically important functional domains = BRCA2 PALB2 binding aa 10-40 and BRCA2 DNA binding aa 2481-3186. Met1090 is outside both.
• Appendix J, Table 16 (BP4 row): 'Apply BP4 for missense or in-frame insertion-deletion inside a (potentially) clinically important functional domain, and no predicted impact via protein change or splicing (BayesDel <= 0.15 (BRCA1) or <= 0.18 (BRCA2), AND SpliceAI <= 0.1).'
• Appendix J, Table 16 (BP1 row): 'Apply BP1_Strong for silent substitution, missense or in-frame insertion-deletion outside a (potentially) clinically important functional domain (no splicing predicted, SpliceAI <= 0.1). Missense prediction not applicable.'
• BayesDel (no-AF) local lookup: -0.499525 - satisfies the BRCA2 no-impact numeric cutpoint (<= 0.18) but cannot satisfy the domain condition of the missense BP4 rule.
• SpliceAI local lookup: max delta 0.00 (<= 0.1) - the numeric splice condition is met, but this splice result is not used as BP4 support for a missense variant.
• REVEL local lookup: 0.319 - indeterminate; generic BP4-supporting requires <= 0.29 (Pejaver et al. 2022, Am J Hum Genet, PMID:36413997).
• Dines et al. 2020 (PMID:31911673) reports BRCA2 exons 10-11 (codons 266-2281, containing codon 1090) as a variation-tolerant 'coldspot' and proposes treating coldspot missense variants as likely benign/under BP4; the paper does not name c.3270G>C, provides no variant-specific data, and is not the ENIGMA-calibrated rule, so it cannot by itself satisfy BP4 here.

Still missing

• Searched the governing VCEP files for this exact variant and found no variant-specific pre-assigned BP4 entry: Specifications_V1.2_2024-11-18.docx, Appendices_V1.2_2024-11-18.docx, PMID_22505045_Houdayer_2012_HumMutat.pdf, SupplementaryTables_V1.2_2024-11-18.xlsx, Specifications_Table9_V1.2_2024-11-18.xlsx, PMID_31853058_BRCA2_clinical_history_LR.xlsx (searched c.3270G>C, p.Met1090Ile, p.M1090I, M1090). BP4 is therefore decided by the VCEP rule.
• Benign bioinformatic evidence for this out-of-domain missense variant is routed by the ENIGMA specification to BP1_Strong, which is outside this group's assigned criteria and must be evaluated by the group owning BP1.

Sources

Reasoning

BP5 in the ENIGMA BRCA2 framework captures combined clinical evidence AGAINST pathogenicity and states 'Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data', with BP5 (supporting) at LR <=0.48:1, BP5_Moderate <=0.23:1, BP5_Strong <=0.05:1 and BP5_VeryStrong <=0.00285:1, and 'Combined LR >0.48-1.00 is not informative (BP5 not applicable)'. This is the likelihood-ratio-style criterion in this group whose operator runs in the opposite direction to PS4/PP4: met requires the value to go DOWN below the threshold. The governing variant-level sources were searched for c.3270G>C, p.Met1090Ile and p.M1090I: the Li et al. 2020 (PMID:31853058) BRCA2 clinical-history LR table has no entry for this variant, the Parsons et al. 2019 SuppT1 combined multifactorial table has no entry, and the Easton et al. 2007 combined multifactorial assessment does not name it. No combined clinical LR therefore exists for this variant. Operator check as required before concluding: the governing operator for BP5 is '<=' (supporting LR <=0.48); with no LR value available, 'LR <= 0.48' cannot be instantiated with a substituted number, so BP5 is not met. The region-level Bayesian reanalysis of ClinVar by PMID:31911673 (BRCA2 exons 10-11 'coldspot', OR <0.01, 95% CI 0.0-0.01, for a region that contains codon 1090) is not a variant-level multifactorial clinical LR, is not specific to this variant, and its authors propose it as region-level benign evidence (a future benign criterion or BP4-type argument) rather than as BP5 clinical evidence, so it is not used to satisfy BP5.

Rule

ENIGMA BRCA2 v1.2 BP5: combined clinical multifactorial LR against pathogenicity - supporting LR <=0.48:1, moderate <=0.23:1, strong <=0.05:1, very strong <=0.00285:1; combined LR >0.48-1.00 not informative.

Evidence considered

• ENIGMA BRCA2 v1.2 BP5 rule text: 'Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data. BP5_VeryStrong - LR <=0.00285:1; BP5_Strong - LR <=0.05:1; BP5_Moderate - LR <=0.23:1; BP5 - LR <=0.48 to >0.23:1. Combined LR >0.48-1.00 is not informative (BP5 not applicable)'.
• The Li et al. 2020 BRCA2 clinical-history LR table and the Parsons et al. 2019 SuppT1 combined multifactorial table were both searched for c.3270G>C, p.Met1090Ile and p.M1090I; neither contains a row for this variant, so no LR can be read or computed.
• PMID:31911673 reports a region-level odds ratio <0.01 (95% CI 0.0-0.01) for BRCA2 exons 10-11 (codons 266-2281, which contain codon 1090), but this is not a variant-level multifactorial clinical LR and the variant is never named; it is not used for BP5.
• Operator direction check for this criterion specifically: the BP5 operator as literally written in the governing ENIGMA text is '<=' (LR <=0.48 supporting), i.e. met requires the value to go DOWN past the threshold - opposite to the '>=' direction used by PS4/PP4 in this same group; no LR value is available to substitute, so the criterion is not met.

Still missing

• No entry for c.3270G>C / p.Met1090Ile / p.M1090I in PMID_31853058_BRCA2_clinical_history_LR.xlsx (the gene-specific PP4/BP5 clinical-history LR table; searched, file available and read).
• No entry for this variant in HUMU-40-1557-s001.xlsx SuppT1 (Parsons et al. 2019, PMID:31131967) combined multifactorial LR table; searched.
• No variant-level combined clinical LR against pathogenicity (family history, pathology, co-occurrence, cosegregation, case-control LR) is available for this variant from any case source.

Sources

Reasoning

BP6 is applied only from an exact-variant ClinVar expert-panel classification of Benign or Likely benign; ordinary laboratory submissions and aggregate ClinVar labels must never trigger it. ClinVar VCV000051441 (variation ID 51441, NM_000059.4(BRCA2):c.3270G>C p.Met1090Ile) does contain two Likely benign submissions (Ambry Genetics, SCV002610893; University of Washington Department of Laboratory Medicine, SCV003851824), but both are single-laboratory assertions with review status 'criteria provided, single submitter'; the remaining four submissions are Uncertain significance (Breast Cancer Information Core, LabCorp Women's Health and Genetics, Color Diagnostics) or a second Labcorp/Invitae submission. The submission audit records expert_panel_submissions = 0, the overall record is 1 star and 'criteria provided, conflicting classifications', and the aggregate API classification is 'Conflicting classifications of pathogenicity'. No ClinGen expert panel (ENIGMA BRCA1/BRCA2 VCEP) classification exists, so no exact-variant expert-panel Benign/Likely benign assertion is present. The ENIGMA BRCA2 v1.2 specification itself lists BP6 as 'Do not use'. BP6 is therefore not met.

Rule

Exact-variant ClinVar expert-panel (ClinGen VCEP) classification of Benign or Likely benign is required to apply BP6; ordinary laboratory submissions and aggregate ClinVar labels do not qualify. ENIGMA BRCA2 v1.2 lists BP6 as 'Do not use'.

Evidence considered

• ClinVar VCV000051441 / variation 51441: Likely benign from Ambry Genetics (2025-04-21) and University of Washington Department of Laboratory Medicine (2023-03-23), both 'criteria provided, single submitter'; Uncertain significance from BIC, LabCorp and Color Diagnostics; expert_panel_submissions = 0; review_stars = 1; aggregate label 'Conflicting classifications of pathogenicity'.
• ENIGMA BRCA2 v1.2 specification lists BP6 as 'Do not use'; per the governing global rule BP6 is nonetheless evaluated for an exact-variant ClinVar expert-panel Benign/Likely benign classification, and none exists.

Still missing

• No ClinGen expert-panel (ENIGMA BRCA1/BRCA2 VCEP) ClinVar classification exists for VCV000051441; the two Likely benign submissions are single-laboratory assertions only.

Sources

Reasoning

ENIGMA BRCA2 v1.2 assigns PP1 for co-segregation of the variant with disease in multiple affected family members of a family, as measured by a quantitative co-segregation method (recommended tools COOL or CAL-Leiden; see Appendix I), with strength set by likelihood ratio: PP1 Supporting LR>=2.08, Moderate LR>=4.3, Strong LR>=18.7, Very Strong LR>=350. A combined LR between 0.48 and 2.08 provides evidence in neither direction, and a score from a single family requires considering an alternative causative pathogenic variant. The case evidence contains no pedigree, no family-member genotype data, and no co-segregation likelihood ratio for c.3270G>C (p.Met1090Ile). The variant was searched for under all three notations (c.3270G>C, p.Met1090Ile, p.M1090I) across the ENIGMA specification, the converted VCEP full texts, HUMU-40-1557-s001.xlsx (Parsons et al. 2019) and SupplementaryTables, and no segregation analysis or family data were returned. Without a quantitative co-segregation LR the PP1 thresholds cannot be evaluated, so PP1 is not assessed.

Rule

ENIGMA BRCA2 v1.2 PP1 by quantitative co-segregation LR: Supporting >=2.08, Moderate >=4.3, Strong >=18.7, Very Strong >=350; neutral 0.48-2.08.

Evidence considered

• ENIGMA BRCA2 v1.2 PP1: co-segregation with disease in multiple affected family members, measured by a quantitative co-segregation method (COOL or CAL-Leiden; Appendix I); Supporting LR>=2.08, Moderate LR>=4.3, Strong LR>=18.7, Very Strong LR>=350; LR 0.48-2.08 provides evidence in neither direction.
• Searched for c.3270G>C, p.Met1090Ile, and p.M1090I across the ENIGMA specification, the converted VCEP full texts, HUMU-40-1557-s001.xlsx and the converted SupplementaryTables file; no segregation data or family information was returned.
• The case evidence provides no pedigree, no genotype of affected relatives, and no segregation likelihood ratio for this variant.

Still missing

• No family pedigree or co-segregation data for c.3270G>C (p.Met1090Ile) is present in the case evidence; a quantitative co-segregation LR (COOL/CAL-Leiden) is required to evaluate PP1.
• No variant-specific PP1 entry for c.3270G>C, p.Met1090Ile, or p.M1090I was found in Specifications_V1.2_2024-11-18.docx or HUMU-40-1557-s001.xlsx.
• The BS4/PS4-governing file SupplementaryTables_V1.2.xlsx declared in the VCEP index could not be converted and its full text is unavailable.

Sources

Reasoning

Governing rule (ENIGMA specification v1.2, Table 1, Table 8, Appendix H): BS2 is not a population-frequency code in this gene framework. It is applied only for a proband with a (presumed) FA-unaffected phenotype in whom the variant under assessment is detected in trans with a P or LP variant in the same gene, or as a homozygote, with points assigned by age at first cancer diagnosis or age at last follow-up (>50y: 4 points in trans / 2 as homozygote / 1 phase-unknown; 40-50y: 2 / 1 / 0.5). BS2 requires >=4 points, BS2_Moderate 2 points, BS2_Supporting 1 point. No proband-level data of this kind (co-occurrence with a classified P/LP BRCA2 variant, phase, zygosity, age at diagnosis or follow-up) exists for this case, so BS2 cannot be evaluated. Population-frequency data alone cannot support BS2 here: Appendix H/Table 13 explicitly states that homozygote-based BS2 points must NOT be assigned from gnomAD-type frequency data, and gnomAD v4.1 reports 0 homozygotes for this variant. BS2 is therefore not assessed rather than not met, because the required proband evidence was never available to the assessment.

Rule

BS2 (variable weight) - Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype: VUA in trans with a P/LP variant in the same gene, or as a homozygote, in a (presumed) FA-unaffected proband, scored by age at diagnosis/follow-up per Table 13 (BS2 >=4 points; BS2_Moderate 2; BS2_Supporting 1). See Appendix H.

Evidence considered

• ENIGMA specification v1.2, BS2: 'Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See Specifications Table 8 for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. BS2 = >=4 points; BS2_Moderate = 2 points; BS2_Supporting = 1 point.'
• Appendices v1.2, Appendix H / Table 13: BS2-related points are assigned per proband in whom the VUA is detected in trans with a P or LP variant in the same gene (or as a homozygote) and who is unlikely to have FA, using age at first cancer onset or age at last follow-up (>50y = 4/2/1 points; 40-50y = 2/1/0.5 points for in-trans/homozygote/phase-unknown respectively).
• Appendices v1.2, Table 13 footnote: 'Apply only for phenotyped individuals from clinical or research cohorts. NOT to be applied for data used to assign frequency-based codes; we have observed that variants observed as >=2 homozygotes in GnomAD data (assumed unaffected adults of unspecified age) are already captured by the BA1 or BS1 frequency codes.'
• gnomAD v4.1 all-comers reports homozygotes = 0 for this variant; no homozygote is available to trigger the BS2 homozygote pathway.
• No proband-level co-occurrence, phase, zygosity, or age-at-diagnosis information for this variant was present in the case evidence.

Still missing

• BS2 requires proband-level data that is absent from this case: a BRCA2 VUS detected in trans with a classified P or LP variant in the same gene (or as a homozygote) in a patient who is FA-unaffected, with age at first cancer diagnosis or age at last follow-up to assign Table 13 points. No such case history, co-occurrence, phase, or zygosity information was available.
• gnomAD v4.1 reports 0 homozygotes for this variant and Appendix H forbids deriving BS2 points from gnomAD-type frequency data, so no BS2 evidence can be extracted from the population resources.
• No VCEP-file entry for this exact variant was found (Specifications_V1.2_2024-11-18.docx, Appendices_V1.2_2024-11-18.docx, HUMU-40-1557-s001.xlsx searched for c.3270G>C / p.Met1090Ile / p.M1090I; no match).

Sources

Reasoning

ENIGMA BRCA2 v1.2 assigns BS4 for lack of co-segregation with disease in affected family members, quantified by a likelihood ratio: BS4 Supporting LR<=0.48, Moderate LR<=0.23, Strong LR<=0.05, Very Strong LR<=0.00285; a combined LR between 0.48 and 2.08 provides evidence in neither direction. The VCEP index declares SupplementaryTables_V1.2.xlsx (ENIGMA posterior-probability/LR table, ST7 reference set) as governing BS4 and PS4, but that file could not be converted and its full text is unavailable. The converted VCEP tables were searched under all three variant notations: no row for c.3270G>C, p.Met1090Ile, or p.M1090I was found in SupplementaryTables_V1.2_2024-11-18.xlsx or HUMU-40-1557-s001.xlsx (Parsons et al. 2019), and the case evidence reports no non-segregation or negative co-segregation analysis. No quantitative BS4 likelihood ratio is available, so BS4 cannot be evaluated and is not assessed.

Rule

ENIGMA BRCA2 v1.2 BS4 by quantitative non-segregation LR: Supporting <=0.48, Moderate <=0.23, Strong <=0.05, Very Strong <=0.00285; neutral 0.48-2.08.

Evidence considered

• ENIGMA BRCA2 v1.2 BS4: lack of co-segregation measured by a quantitative method; Supporting LR<=0.48, Moderate LR<=0.23, Strong LR<=0.05, Very Strong LR<=0.00285; LR 0.48-2.08 provides evidence in neither direction.
• The VCEP index declares SupplementaryTables_V1.2.xlsx (ENIGMA posterior probability table) as governing BS4 and PS4; this file is unavailable because it could not be converted to text.
• Searched for c.3270G>C, p.Met1090Ile, and p.M1090I in the converted SupplementaryTables_V1.2_2024-11-18.xlsx and HUMU-40-1557-s001.xlsx; no matching variant row was found.
• The case evidence reports no lack-of-segregation or negative co-segregation analysis for this variant.

Still missing

• The VCEP-index-declared file SupplementaryTables_V1.2.xlsx (ENIGMA posterior probability / LR table) is not available in converted form (could not be converted), so any pre-assigned BS4 code for p.Met1090Ile could not be retrieved from it.
• No variant-specific BS4 entry for c.3270G>C, p.Met1090Ile, or p.M1090I was found in the converted SupplementaryTables_V1.2_2024-11-18.xlsx or in HUMU-40-1557-s001.xlsx.
• The case evidence contains no lack-of-segregation analysis or co-segregation likelihood ratio; a quantitative non-segregation LR is required to evaluate BS4.

Sources