Final classification

Pathogenic

PVS1PS3PM2PP3

RAD51C

c.904+5G>T

p.?

unknown · exon 6i

Gene context NCBI Gene ↗ OncoKB ↗

RAD51C encodes a member of the RAD51 recombinase family that is central to repairing double-stranded DNA breaks via homologous recombination, helping assemble RAD51 filaments and resolve Holliday junctions. Mutations in this gene cause Fanconi anemia complementation group O, and germline mutations increase susceptibility to ovarian cancer and possibly breast cancer. RAD51C acts as a tumor suppressor: it is infrequently mutated in human cancers, but its expression is reduced in a subset of breast tumors, and it resides in a region of chromosome 17 that is frequently amplified in breast cancer.

This variant

RAD51C maintains homologous-recombination repair of double-stranded DNA breaks, and loss of that function causes Fanconi anemia group O while germline loss-of-function variants raise ovarian and breast cancer susceptibility. Classifying c.904+5G>T as Pathogenic places it among those loss-of-function alleles, since it removes exon 6 and is predicted to trigger nonsense-mediated decay of the transcript. For this patient, the finding indicates an established mechanism of RAD51C dysfunction linked to hereditary cancer risk.

Transcript

NM_058216.3

HGVS · transcript:coding

NM_058216.3:c.904+5G>T

GRCh38

chr17:58720817 G>T

GRCh37

chr17:56798178 G>T

Pathogenic: PVS1 (Very Strong) combined with PS3 (Strong) met the generic ACMG/AMP combination rule, with PM2 and PP3 providing supporting evidence.

Classification rationale

PVS1PS3PM2PP3 Pathogenic

RAD51C c.904+5G>T unknown · exon 6i

PVS1 (Very Strong): exon 6 exclusion produces the frameshift V280GfsX11, predicted to trigger nonsense-mediated decay. PS3 (Strong): patient tumor RNA from two carriers and an independent minigene assay both reproduced variant-caused exon 6 skipping. PM2 (Supporting): absent or ultra-rare in population databases, with gnomAD v4.1 frequency 0.000873% and no homozygotes. PP3 (Supporting): SpliceAI maximum delta 0.949 predicts a splice impact, above the >0.2 threshold. Overall classification: Pathogenic, from PVS1 (Very Strong) plus PS3 (Strong) under the generic ACMG/AMP combination rule.

PVS1 + PS3 + PM2 + PP3 → Pathogenic

LYFE Sciences is an AI system, and it can make mistakes. Criteria may be applied incorrectly, sources may be misread, and a confident-looking classification can still be wrong. Double-check every criterion and its underlying evidence before relying on any call.

Gene diagram · NM_058216.3 · variants mapped to exon structure

RAD51C NM_058216.3

Fetching transcript structure from UCSC…

Applied criteria · 4 applied · 16 assessed

MetEvidence satisfies this criterion.

Not metEvaluated against available evidence; threshold not reached.

Not assessedApplies in principle, but no evidence was found to evaluate it.

N/ADoesn't apply to this variant type.

Applied · 4

Strength Supporting Moderate Strong Very strong

PVS1 very strong Pathogenic

Met (Very Strong): c.904+5G>T causes exon 6 exclusion, producing the frameshift V280GfsX11 predicted to trigger nonsense-mediated decay.

The applicable RAD51C CSPEC identifies NM_058216.3 as the preferred transcript but contains no structured PVS1 rule; generic ClinGen SVI PVS1 guidance was therefore used for this criterion.PMID:20400964 reports c.904+5G>T as IVS 6 904+5G>T with predicted consequence V280GfsX11. RT-PCR from two carrier tumor samples yielded a product lacking exon 6, and a RAD51C exon-6 splicing reporter independently demonstrated exon 6 exclusion.The PVS1 gene-context record supports RAD51C loss of function as a germline disease mechanism and permits generic PVS1 assessment.

cspec ↗pvs1_generic_framework ↗PMID:20400964 ↗

PS3 strong Pathogenic

Met (Strong): patient tumor RNA from two carriers and an independent minigene reporter both reproduced variant-caused exon 6 exclusion.

The RAD51C VCEP specification is applicable to this case, but its extracted criterion payload is empty and does not provide a PS3-specific strength calibration; therefore the assay evidence was calibrated using generic ACMG/AMP principles.PMID:20400964 reports the exact 904+5G>T variant in two hereditary breast/ovarian-cancer pedigrees. RT-PCR of paraffin-embedded tumor RNA from two carriers produced an exon-6-lacking transcript that was absent from controls.PMID:20400964 independently tested the exact variant in a RAD51C exon-6 splicing reporter containing flanking intronic sequence and found variant-associated exon-6 exclusion, directly supporting loss of normal splicing.

cspec ↗PMID:20400964 ↗

PM2 supporting Pathogenic

Met (Supporting): ultra-rare in population databases at gnomAD v4.1 frequency 0.000873% (14/1,602,834 alleles), with no homozygotes observed.

The RAD51C ClinGen CSPEC record is present but marked incomplete with no usable PM2 population rule; generic ACMG/AMP guidance was therefore applied.Generic ACMG/AMP PM2 supports a pathogenic interpretation when a variant is absent or extremely rare in population databases, accounting for database quality and disease prevalence (Richards et al., PMID:25741868).gnomAD v4.1 reports 14/1,602,834 alleles, AF 8.73453e-06, and zero homozygotes; the highest broad population AF is 1.33955e-05.

cspec ↗gnomad_v2 ↗gnomad_v4 ↗gnomad_canada ↗

PP3 supporting Pathogenic

Met (Supporting): SpliceAI maximum delta 0.949 exceeds the >0.2 threshold, predicting a splice impact on this non-canonical donor-region variant.

The supplied RAD51C CSPEC record identifies the governing framework but contains an empty criteria payload and no PP3/BP4-specific rule or lookup assignment.SpliceAI Lookup reports DS_AL=0.949, DS_DL=0.940, and a maximum delta score of 0.949 for NM_058216.3:c.904+5G>T. Applying the generic non-missense SpliceAI calibration (PP3 when maximum delta >0.2) yields PP3 at supporting strength.PMID:34299313 independently reports that c.904+5G>T was predicted to inactivate the RAD51C intron-6 wild-type donor site, but this paper-specific Human Splicing Factor prediction was not used to add or increase PP3 strength; the calibrated SpliceAI result alone supports this call.

cspec ↗spliceai ↗

Assessed · not applied · 5 not met · 11 not assessed

Pathogenic

PS2 Not assessed: the exact variant appears in hereditary breast and ovarian cancer pedigrees, but no parental genotypes or confirmed de novo occurrence are documented.

PS4 Not assessed: the variant appears in affected breast and ovarian cancer cohorts, but no case-control enrichment statistic or specified counts were available.

PM3 Not assessed: reported in breast and ovarian cancer pedigrees, but no confirmed biallelic, in-trans, or compound-heterozygous Fanconi anemia case establishes PM3.

PM6 Not assessed: reports describe inherited cancer pedigrees without parental genotypes or a confirmed de novo occurrence for this variant.

PP1 Not assessed: two cancer pedigrees are reported, but no countable meioses, relative genotypes, or segregation statistic were available.

PP4 Not assessed: breast cancer and ovarian serous carcinoma fit RAD51C-associated contexts but are not a highly specific single-gene phenotype.

PP5 Not met: ClinVar holds only laboratory, research, and curation submissions, with no expert-panel pathogenic assertion for this exact variant.

Benign

BA1 Not met: gnomAD v4.1 frequency of 0.000873% (14/1,602,834 alleles) is far below the >5% stand-alone benign population threshold.

BS1 Not met: the observed population frequency is extremely low and shows no excess over that expected for a rare RAD51C disease-associated variant.

BS2 Not assessed: population databases show zero homozygotes but no confirmed healthy adult carrier, so BS2 cannot be evaluated.

BS3 Not assessed: the variant-specific assays demonstrate abnormal splicing rather than preserved normal function, so no benign functional evidence exists.

BS4 Not assessed: no tested unaffected relatives lacking the variant and no family-level non-segregation analysis were available.

BP2 Not assessed: reports do not establish the variant in trans with a pathogenic allele, or in cis in a dominant disease context.

BP4 Not met: BP4 requires a SpliceAI maximum delta below 0.1, but the observed delta is 0.949.

BP5 Not assessed: a coexisting non-coding ATM variant was reported in one ovarian cancer patient without established pathogenicity, so no alternate diagnosis is confirmed.

BP6 Not met: ClinVar contains no expert-panel benign or likely benign assertion for this exact variant.

N/A · 8 PS1 · PM1 · PM4 · PM5 · PP2 · BP1 · BP3 · BP7

Research & evidence

Population frequency · supports pathogenic

v4.1

This variant is present in gnomAD v4.1 (AF= 8.73453e-06; MAF= 0.00087%, 14/1602834 alleles, homozygotes = 0) and has highest observed frequency in the African/African American population (AF= 1.33955e-05; MAF= 0.00134%, 1/74652 alleles, homozygotes = 0); grpmax FAF= 6.19e-06.

v2.1

This variant is present in gnomAD v2.1 (AF= 1.59658e-05; MAF= 0.00160%, 4/250536 alleles, homozygotes = 0) and has highest observed frequency in the European (non-Finnish) population (AF= 3.52361e-05; MAF= 0.00352%, 4/113520 alleles, homozygotes = 0); grpmax FAF= 1.123e-05.

🇨🇦 CA

Absent from gnomAD-Canada v1.0.

Allele frequency by ancestry

three datasets · side by side

gnomAD v4.1

0.00087% · 14 / 1,602,834
0 hom · FAF 0.00062%

African/African American 1 / 74,652	0.0013%
European (non-Finnish) 13 / 1,171,360	0.0011%

+ 8 not observed (Remaining individuals, Admixed American, European (Finnish), Amish, East Asian, Middle Eastern, South Asian, Ashkenazi Jewish)

gnomAD v2.1

0.0016% · 4 / 250,536
0 hom · FAF 0.0011%

European (non-Finnish)

4 / 113,520

0.0035%

+ 7 not observed (African/African American, Admixed American, Ashkenazi Jewish, East Asian, European (Finnish), Remaining individuals, South Asian)

gnomAD Canada 🇨🇦

Absent · 0 / ?
0 hom

Not observed in any ancestry group.

gnomAD v4 ↗ gnomAD v2 ↗ gnomAD 🇨🇦 ↗

ClinVar

This variant has been reported in ClinVar as Likely pathogenic (15 clinical laboratories) and as Pathogenic (4 clinical laboratories) and as pathogenic (1 clinical laboratory). (ClinVarID = 142762)

ClinVar ↗

In silico

SpliceAI predicts possible splice impact for this variant (max delta score = 0.95).

SpliceAI ↗

Functional No data

No calibrated functional assay or RNA evidence was identified for this variant.

OncoKB ↗

Somatic evidence

COSMIC

This variant has not previously been reported in somatic cancers (COSMIC).

Hotspots

This variant does not lie in a statistically significant cancer hotspot.

COSMIC ↗

Literature · how each cited paper was used

2papers cited

Each card is an audit: what was searched, what was found, whether it names the variant, which criteria it fed, and why. 6 further PMIDs triaged but not cited — see Sources & references.

Germline mutations in breast and ovarian cancer pedigrees establish RAD51C as a human cancer susceptibility gene.

PMID 20400964 ↗ · Meindl A et al. · 2010 May CLINVAR · functional

Searched

c.904+5G>T904+5G>TNP_478123.1:p.?V280GfsX11

Found

The paper explicitly reports RAD51C c.904+5G>T, listed as IVS 6 904+5G>T with the predicted protein consequence V280GfsX11. The variant was identified in two breast/ovarian cancer pedigrees within the cohort of 1,100 German hereditary gynecological-cancer families. RNA analysis from paraffin-embedded tumor samples showed an RT-PCR product lacking exon 6 in carriers, and a RAD51C exon-6 splicing reporter assay showed that the variant caused exon 6 exclusion. Loss of the wild-type allele was observed independently in breast and ovarian tumor tissues from two affected individuals in the relevant pedigree.

Variant

✓ Names this variant — characterised directly

Applied to

→PVS1 very strong

Variant-specific RNA and minigene results establish exon 6 exclusion, with the reported frameshift consequence V280GfsX11.

→PS3 strong

The exact variant causes exon 6 exclusion in patient RNA and in an independent splicing reporter assay, with control comparison.

→PP3 supporting

Variant-specific prediction of impaired donor recognition was reviewed as concordant context; the calibrated SpliceAI result, not this prediction, determines the PP3 call.

The second splice site mutation (904+5G>T) also affected an evolutionarily conserved position and was predicted to greatly decrease the complementarity between the U1 small nuclear RNA and the 5′ splice site: the HBond score11 decreased from 15.8 to 10.1, a predictor of aberrant splicing. As cells carrying the germline mutation were not available (Fig. 1d), mRNA was isolated from paraffin-embedded tumor samples from two carriers of this mutation. As shown in Figure 2e, we were able to specifically amplify an RT-PCR product that lacked exon 6 in these carriers (Supplementary Fig. 1b) but not in control samples. For further confirmation, we introduced the RAD51C exon 6 with flanking 225- and 158-bp intronic sequences into a splicing reporter construct12 (Fig. 2f). Functional analysis in this heterologous context also demonstrated that the 904+5G>T mutation resulted in exclusion of exon 6, suggesting that the RAD51C exon 6 is recognized by exon definition (Fig. 2g).

Location Table 1; Results, Nature Genetics vol. 42, p. 412, paragraph describing the second splice-site mutation; Figure 2e–g; Figure 1d · Context Screening of 1,100 German hereditary breast/ovarian-cancer pedigrees; RT-PCR of RNA from paraffin-embedded tumor samples from two carriers; RAD51C exon 6 minigene/splicing reporter assay in a heterologous system; loss-of-heterozygosity analysis of breast and ovarian tumor tissues. · full text

Detection of Germline Variants in 450 Breast/Ovarian Cancer Families with a Multi-Gene Panel Including Coding and Regulatory Regions.

PMID 34299313 ↗ · Guglielmi C et al. · 2021 Jul 19 CLINVAR · functional

Searched

c.904+5G>Tc.904+5G > TNP_478123.1:p.?

Found

This paper explicitly reports RAD51C c.904+5G>T as an intron 6 splice-region variant. Human Splicing Factor analysis predicted inactivation of the wild-type donor site; the paper lists the variant in affected patients with breast or ovarian cancer. The report contains no residue-level missense consequence or functional domain assessment.

Variant

✓ Names this variant — characterised directly

Applied to

→PP3 supporting

Variant-specific donor-site prediction was reviewed as concordant context; the calibrated SpliceAI result, not this prediction, determines the PP3 call.

In splicing regions, we identified six variants and only one, c.904+5G > T, was predicted to inactivate the wild-type donor site in the intron 6 of the RAD51C gene, and therefore is classified by the ACMG criteria as ‘Pathogenic’.

Location Results, Section 2.4, paragraph on splicing regions; Table 6, patient Pisa 1564; Table 7, patient Varese 43 · Context Germline targeted NGS panel analysis of 450 Italian breast/ovarian cancer patients; Illumina MiSeq sequencing of coding regions and 50 bp flanking intron-exon junctions; Sanger confirmation; Human Splicing Factor computational prediction. · full text

Sources & reference links

8Sources

CSpec VCEP

ClinVar

gnomAD v2.1

gnomAD v4.1

gnomAD-Canada

SpliceAI

OncoKB

COSMIC

Triaged references · 6 PMIDs not cited in assessment

17576681 ↗ Aberrant 5' splice sites in human disease genes: mutation pattern, nucleotide structure and comparison of computational tools that predict their utilization. CLINVAR

25525159 ↗ RNA splicing. The human splicing code reveals new insights into the genetic determinants of disease. CLINVAR

25741868 ↗ Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. CLINVAR

29758562 ↗ Clinical relevance of systematic phenotyping and exome sequencing in patients with short stature. CLINVAR

31300551 ↗ One in three highly selected Greek patients with breast cancer carries a loss-of-function variant in a cancer susceptibility gene. CLINVAR

31843900 ↗ Characterization of splice-altering mutations in inherited predisposition to cancer. CLINVAR

Check your email

Reset your password

Enter an access code

Your allowance