NTHL1 c.244C>T, p.(Gln82Ter)

Reasoning

NM_002528.7 is the MANE Select / RefSeq Select transcript for NTHL1 (CCDS10457.2; CDS c.1-915 encoding a 305-residue protein), and the variant is a nonsense change predicting NP_002519.2:p.(Gln82Ter). Step 1 of the ClinGen SVI PVS1 framework (PMC6185798) is satisfied: this is a predicted null variant (nonsense, premature termination codon at codon 82). Step 2 is satisfied: loss of function is the established disease mechanism for NTHL1 - biallelic truncating NTHL1 alleles cause NTHL1 tumor syndrome / adenomatous polyposis and colorectal cancer, and the variant-specific report PMID:37727376 identifies c.244C>T p.(Gln82Ter) (originally c.268C>T p.(Gln90Ter) under NM_002528.6, ClinVar 192319) as the recurrent truncating NTHL1 allele, seen in homozygosity in four polyposis patients and in compound heterozygosity in two others, whose tumours show the C>T transition / COSMIC SBS30 mutational signature of defective base excision repair. Step 3 (NMD escape) is not triggered: c.244 lies in exon 2 of 6 (c.116-354), not in the last exon (c.792-*104, exon 6) and not within the last 50 bp of the penultimate exon, and the PTC sits roughly 108 nt upstream of the exon 2/3 junction (c.354/355), so the transcript is predicted to undergo NMD. The truncated/altered region is critical and large: the stop at codon 82 removes the carboxy-terminal 224 of 305 residues (~73%), far beyond the >10% protein-removal threshold, and eliminates essentially the entire glycosylase catalytic region. An NMD-competent null that removes a critical region and >10% of the protein, in a gene where LoF causes disease, is PVS1 at full, un-downgraded strength, i.e. very strong. No named SVI or VCEP downgrade rule applies - the variant does not escape NMD, the affected exon is present in the biologically relevant MANE Select transcript, and the truncation is not confined to a non-critical distal region. SpliceAI (max delta 0.18, below the 0.2 supporting cut-off) adds no independent effect; the primary consequence remains the premature termination codon. Caveats: NTHL1-related disease is autosomal recessive, so PVS1 establishes only that this is a genuine null allele and disease expression requires a second pathogenic allele in trans; and c.244C>T is a common European founder allele (gnomAD v4.1 NFE AF ~0.18%, with one homozygote reported), which does not downgrade PVS1 but limits the standalone weight of a single allele.

Rule

ClinGen SVI PVS1 decision tree (Abou Tayoun et al. 2018, PMC6185798), generic fallback: predicted null variant in a gene where loss of function is the established disease mechanism, NMD-competent (not in the last exon or last 50 bp of the penultimate exon), truncating region critical and removing >10% of the protein -> PVS1 very strong (no downgrade applied).

Evidence considered

• Variant normalization (Mutalyzer / VariantValidator): NM_002528.7:c.244C>T is a single-nucleotide substitution predicting NP_002519.2:p.(Gln82Ter); NM_002528.7 is annotated MANE Select and RefSeq Select (CCDS10457.2), and the reference protein is 305 residues (position_last_original 305, predicted truncation product ending at position 82).
• Transcript/exon mapping: c.244 falls in exon 2 of 6 (exon 2 = c.116-354); the last exon is c.792-*104 (exon 6) and the CDS spans c.1-915. The variant is therefore neither in the last exon nor in the last 50 bp of the penultimate exon, and the premature stop lies ~108 nt upstream of the exon 2/3 junction (c.354/355), consistent with NMD-competent rather than NMD-escaping.
• Protein-length change: the codon-82 stop removes residues 82-305, i.e. 224 of 305 residues (~73%) of the protein, which is far above the >10% threshold used by the ClinGen SVI PVS1 framework for retaining full (very strong) strength.
• Loss-of-function mechanism: no NTHL1-specific ClinGen CSPEC or VCEP specification is available, so the generic ClinGen SVI PVS1 framework (PMC6185798) governs; the gene-level review recorded for this case concludes that germline NTHL1 loss of function is a known disease mechanism, making the framework's step-2 gate pass (PVS1 eligible).
• PMID:37727376 (variant-specific): reports NM_002528.7:c.244C>T p.(Gln82Ter) as the recurrent truncating NTHL1 variant underlying NTHL1 tumor syndrome (previously reported as c.268C>T p.(Gln90Ter) under NM_002528.6, ClinVar 192319), found homozygous in four polyposis patients and compound heterozygous in two more, with tumours showing non-CpG C>T transitions and the COSMIC SBS30 signature, consistent with defective base excision repair. This supports both the null consequence and the LoF disease mechanism.
• PMID:25741868 (framework): the ACMG/AMP 2015 standards define PVS1 as applying to null variants (nonsense, frameshift, canonical +/-1 or 2 splice sites, initiation codon, single- or multi-exon deletion) that can be assumed to abolish gene function, subject to the caveats that null variants are an established disease mechanism and that variants in the last exon or the last 50 bp of the penultimate exon (NMD escape) be treated differently; neither caveat applies here.
• SpliceAI (max delta 0.18; DS_DG 0.183) is below the 0.2 supporting cut-off and indicates no significant additional splice effect, so the null consequence assigned to this nonsense allele stands unchanged.
• No SVI or VCEP downgrade rule was triggered: the transcript is the clinically relevant MANE Select isoform, the affected exon is not subject to a documented exclusion, and the truncation removes a critical region rather than a non-critical distal segment.

Still missing

• No ClinGen CSPEC or VCEP specification exists for NTHL1, so no gene-specific PVS1 carve-out or transcript list was available; the generic ClinGen SVI PVS1 decision tree (PMC6185798) was applied.
• Zygosity and phase are not established in the available evidence: NTHL1 disease is autosomal recessive, so PVS1 confirms the null nature of this allele but a second pathogenic allele in trans is required for disease expression.
• Protein-domain boundaries for NTHL1 (helix-hairpin-helix / [4Fe-4S] catalytic region) were inferred from the truncation size and the recessive LoF mechanism rather than from a supplied domain table.

Sources

Reasoning

PS3 is met at supporting strength. The case variant NM_002528.7:c.244C>T p.(Gln82Ter) is an NTHL1 truncating allele; in the original NTHL1 report (PMID:25938944) this exact variant appears under the legacy NM_002528.6 name c.268C>T p.(Gln90*), and PMID:37727376 explicitly establishes that c.268C>T p.(Gln90Ter) (NM_002528.6) and c.244C>T p.(Gln82Ter) (NM_002528.7, ClinVar 192319) are the same allele. Weren et al. performed a variant-specific RNA assay on this allele: EBV-transformed B lymphocytes from homozygous carriers were analysed by real-time quantitative RT-PCR, with cycloheximide treatment to block nonsense-mediated decay as an internal assay control, and the authors state the nonsense mutation 'results in nonsense-mediated decay.' A clinical submission in ClinVar (PreventionGenetics, SCV004116067) summarizes the same assay as demonstrating nonsense-mediated decay with a ten-fold reduction in NTHL1 RNA versus controls. This is a variant-specific, patient-cell functional readout with a built-in NMD control (cycloheximide) showing a deleterious effect on the gene product (loss of NTHL1 transcript). Independent in-vivo corroboration comes from PMID:37727376, in which tumours from homozygous carriers of this exact variant show a predominance of C>T transitions at non-CpG sites and the COSMIC SBS30 mutational signature, the expected functional consequence of defective NTHL1 base-excision repair. Strength is calibrated to supporting rather than moderate/strong for three reasons: (i) the readout is transcript abundance / NMD, which is the expected consequence of a premature stop codon and is already captured by PVS1 for this truncating allele, so attributing strong independent functional weight would double-count the same loss-of-function biology; (ii) no published, assay-specific PS3 calibration (e.g. ClinGen SVI PS3/BS3 validation) exists for this qRT-PCR/NMD assay, so it cannot be graded above supporting on validated performance; and (iii) at least one clinical submission (SCV005902063, Catalan Institute of Oncology) states 'to our knowledge, no well-established functional studies have been reported for this variant,' showing the functional-evidence base is not uncontested. needs_human_review is set because the supporting-versus-moderate strength assignment and the PVS1 overlap require adjudication.

Rule

ACMG/AMP 2015 PS3 (PMID:25741868): well-established functional studies show a deleterious effect. Generic fallback (no NTHL1 VCEP/CSPEC). Applied at supporting strength: variant-specific RNA/NMD assay in patient-derived B lymphocytes demonstrates loss of the NTHL1 transcript, but the readout is transcript-level and overlaps the PVS1 loss-of-function mechanism, and no published assay-specific calibration supports a higher strength.

Evidence considered

• PMID:25938944 (Weren et al. 2015), Online Methods 'RNA expression analysis': EBV-transformed B lymphocytes from homozygous carriers were cultured, treated with cycloheximide for 4 h to block nonsense-mediated decay, and analysed by real-time quantitative RT-PCR. In the main text the authors state that the NTHL1 nonsense mutation encoding p.Gln90* (c.268C>T; NM_002528) 'results in nonsense-mediated decay.' This is the variant-specific functional assay that supports PS3.
• PMID:37727376 establishes the nomenclature bridge: the Weren 2015 allele reported as c.268C>T p.(Gln90Ter) under NM_002528.6 is the same allele as the case variant c.244C>T p.(Gln82Ter) under NM_002528.7 (ClinVar 192319), so the Weren functional data are variant-specific for NM_002528.7:c.244C>T.
• ClinVar submission SCV004116067 (PreventionGenetics, part of Exact Sciences) states: 'functional studies in b-lymphocytes from homozygous carriers indicate that this variant results in nonsense mediated decay and a ten-fold reduction in nthl1 rna expression in comparison to controls'. (The stored sentence is truncated after '(w' in the source record.)
• ClinVar submission SCV005902063 (Molecular Diagnostics Laboratory, Catalan Institute of Oncology) states 'to our knowledge, no well-established functional studies have been reported for this variant', which is recorded as a counterpoint on whether an NMD assay qualifies as a 'well-established' functional study.
• ClinVar submission SCV000992218 (University of Cambridge) lists PS3 with the phrase 'loss of heterozygosity in tumours or immunohistochemistry abnormalities considered functional evidence of pathogenicity'; this is a submitter's generic criterion template and is not variant-specific functional data, so it was not weighted.
• PMID:37727376 (variant-specific, in vivo corroboration): tumours from homozygous c.244C>T p.(Gln82Ter) carriers showed a predominance of C>T transitions at non-CpG sites and the COSMIC SBS30 mutational signature, the functional hallmark of defective NTHL1 base-excision repair.
• PMID:25741868 (ACMG/AMP 2015): PS3 is defined as 'well-established functional studies show a deleterious effect', the definition against which this assay was judged.
• Strength calibration: no gene-specific VCEP/CSPEC specification exists for NTHL1 (vcep_materials records none), so generic ACMG/AMP 2015 governs; the transcript-level NMD assay was graded supporting rather than strong to avoid double-counting the loss-of-function mechanism already captured by PVS1 and because no published assay-specific PS3 calibration is available.

Still missing

• No published ClinGen SVI / assay-specific calibration exists for this B-lymphocyte qRT-PCR nonsense-mediated-decay assay, so PS3 strength cannot be graded above supporting on validated assay performance.
• No direct enzymatic NTHL1 DNA-glycosylase activity assay on the variant protein has been reported; the functional readout is transcript-level (NMD), not protein-function-level.
• The functional data for this allele are reported under the legacy NM_002528.6 name c.268C>T p.(Gln90*); the equivalence to NM_002528.7:c.244C>T p.(Gln82Ter) rests on the nomenclature bridge in PMID:37727376 and should be confirmed against the primary submission.
• PS3 overlaps the loss-of-function biology already captured by PVS1 for this truncating allele; the final classification should avoid double-counting these two evidence lines.

Sources

Reasoning

Met. No NTHL1 ClinGen CSPEC or VCEP exists for this case (cspec found = false, gene_vcep_dir = null, framework_mode = generic_acmg), so PS4 is governed by the generic ACMG/AMP 2015 definition: the prevalence of the variant in affected individuals is significantly increased compared with controls. PS4 in this gene therefore has no VCEP-supplied numeric cut-off; its own governing definition is a significance-based enrichment rule whose direction strengthens upward (odds ratio > 1 with a confidence interval excluding 1). Because no numeric operator exists in any VCEP source text for this criterion, the comparison is taken from PS4's own generic definition and not from any neighbouring criterion: explicit inequality 44.7 > 1? yes (95% CI 6.90-290 excludes 1; P = 6.7 x 10^-5 < 0.05) -> met. The variant-specific evidence is PMID:38036545 (Nurmi et al. 2023, Sci Rep), which evaluated all coding variants in 18,786 breast-cancer cases and 182,927 controls from FinnGen and reported that NTHL1 c.244C>T p.(Gln82Ter) associated with breast cancer with a high risk in homozygous carriers (OR = 44.7, 95% CI 6.90-290, P = 6.7 x 10^-5) and a lower but still significant risk in heterozygous women (OR = 1.39, 95% CI 1.18-1.64, P = 7.8 x 10^-5); the same analysis suggested a high risk of colorectal, urinary-tract and basal-cell skin cancer in homozygotes. This is a direct, variant-specific case-control enrichment measured in an independent population cohort, and the biallelic (homozygous) comparison is the appropriate statistic for this autosomal recessive condition. PMID:37727376 adds independent case-series corroboration of affected-individual occurrence: the same allele (reported there as c.244C>T p.(Gln82Ter), NM_002528.7) was found in homozygosity in four unrelated polyposis patients and in compound heterozygosity with a second truncating/likely pathogenic NTHL1 variant in two further patients, although that series carries no control comparison and therefore contributes case accrual rather than formal enrichment evidence. PMID:18515411 is a case-control screen of NTH1 in 134 multiple-adenoma patients versus ~359 controls; it reports no enrichment for the NTH1 nonsense allele it detected (1/134 cases vs 2/359 controls) but does not name this nucleotide change, so it neither supports nor refutes PS4 for c.244C>T. Strength is left at the generic default (Strong) because no gene-specific PS4 calibration or VCEP threshold table exists for NTHL1; reviewer may consider Moderate if the case-control signal is judged to be phenotype-limited (the primary association is breast cancer rather than the classic polyposis/CRC phenotype, although colorectal and other tumours were also enriched).

Rule

PS4 (generic ACMG/AMP 2015, PMID:25741868): prevalence of the variant in affected individuals significantly increased vs controls; no NTHL1 VCEP numeric cut-off exists, so the default strength (Strong) applies.

Evidence considered

• PMID:38036545 ('NTHL1 is a recessive cancer susceptibility gene', Nurmi et al. 2023, Scientific Reports): genotyping of candidate variants in 2482-4101 breast-cancer patients vs 1273-3985 controls, plus all coding variants in 18,786 breast-cancer cases and 182,927 controls from FinnGen. Variant-specific result: NTHL1 c.244C>T p.(Gln82Ter), homozygous OR = 44.7 (95% CI 6.90-290, P = 6.7 x 10^-5) and heterozygous OR = 1.39 (95% CI 1.18-1.64, P = 7.8 x 10^-5) for breast cancer, with a suggested high risk of colorectal, urinary-tract and basal-cell skin cancer in homozygotes.
• Direction check for PS4's own operator: generic ACMG/AMP 2015 PS4 requires prevalence in affecteds to be significantly increased versus controls (enrichment criterion, strengthens as the value rises). No VCEP source text defines a numeric threshold for PS4 in NTHL1 (no CSPEC/VCEP available), so the operator is taken from PS4's own definition only. Explicit inequality as written: '44.7 > 1?' yes; 95% CI 6.90-290 excludes 1; 'P = 6.7 x 10^-5 < 0.05?' yes -> met.
• PMID:37727376: same allele (c.244C>T p.(Gln82Ter) under NM_002528.7, originally reported as c.268C>T p.(Gln90Ter) under NM_002528.6, ClinVar ID 192319) found in homozygosity in four polyposis patients and in compound heterozygosity with c.435_446del p.(Leu146_Arg149del) (patient #5) or c.503T>C p.(Ile168Thr) (patient #6); homozygous carriers presented colorectal polyposis (<10 to >100 polyps) and colorectal cancer at ages 47-63. Provides case accrual for the same variant; no control group, so not formal case-control enrichment.
• PMID:18515411: NTH1 (NTHL1) screened in 134 patients with multiple colorectal adenomas vs ~359 British controls; the single NTH1 nonsense allele detected showed no enrichment (1/134 cases vs 2/359 controls) and the authors concluded NTH1 does not contribute significantly to autosomal recessive polyposis. The c.244C>T nucleotide change is not named in this paper, so it neither supports nor refutes PS4 here.
• Co-occurring ClinVar condition labels for VCV000192319 (colorectal cancer, familial adenomatous polyposis 3, NTHL1-deficiency tumour predisposition syndrome) were noted but were not used as enrichment evidence.

Still missing

• No NTHL1 VCEP/ClinGen CSPEC calibration exists, so PS4 strength cannot be gene-specifically calibrated; the generic default (Strong) was used.

Sources

Reasoning

NTHL1-associated disease (NTHL1 tumor syndrome / familial adenomatous polyposis 3) is an autosomal-recessive disorder, and c.244C>T p.(Gln82Ter) is a recurrent loss-of-function allele of this gene. The variant has been documented in trans with a second (likely) pathogenic NTHL1 allele: in PMID:37727376 patient #5 it occurs in compound heterozygosity with c.435_446del p.(Leu146_Arg149del) (classified there as likely pathogenic via PM2+PP1+PM4+PM3), and family studies in that pedigree confirmed the biallelic/trans configuration (relatives carrying only one of the two variants). Four additional patients (#1-#4) in the same series were homozygous for c.244C>T with polyposis/CRC, and patient #6 carried it in trans with the VUS c.503T>C p.(Ile168Thr). The founding report (PMID:25938944) identified the same allele — reported then as c.268C>T p.(Gln90*) under NM_002528.6, which corresponds to c.244C>T p.(Gln82*) under NM_002528.7 / ClinVar 192319 — in the homozygous state in seven affected individuals from three unrelated families with recessive inheritance of adenomatous polyposis, with the allele found only heterozygously in controls. A ClinVar submission additionally records the allele compound heterozygous with a canonical splice variant, confirmed in trans by familial testing. Trans/biallelic occurrence in an autosomal-recessive gene therefore meets PM3. Strength is left at the generic default Moderate; it was not upgraded because the best-characterized trans partner is classified likely pathogenic (not pathogenic) and the alternative trans partner is a VUS.

Rule

PM3 (Pathogenic, Moderate): For recessive disorders, detected in trans with a pathogenic variant (Richards et al. 2015, PMID:25741868).

Evidence considered

• NTHL1 tumor syndrome (also described as familial adenomatous polyposis 3 / NTHL1-deficiency tumor predisposition syndrome) is an autosomal-recessive disorder; the ClinVar record for this variant is submitted under these recessive conditions.
• PMID:37727376, Results 3.1 and Table 1: c.244C>T p.(Gln82Ter) was found in homozygosity in polyposis/CRC patients #1-#4, in compound heterozygosity in patient #5 with c.435_446del p.(Leu146_Arg149del), and in compound heterozygosity in patient #6 with c.503T>C p.(Ile168Thr). Direct quote (patient #5): 'Patient #5 was a compound heterozygous carrier presenting the recurrent variant NM_002528.7:c.244C>T p.(Gln82Ter) in concomitance with the NM_002528.7:c.435_446del p.(Leu146_Arg149del) variant.'
• PMID:37727376, family analysis of patient #5's family (Figure 1C): trans phase is confirmed by relatives who carry only one of the two variants — 'The fact that there are relatives harboring only one of the variants confirms the biallelic nature of these variants.' The companion variant c.435_446del is classified by the authors as likely pathogenic (PM2+PP1+PM4+PM3); the c.503T>C p.(Ile168Thr) partner in patient #6 is classified as a VUS.
• PMID:25938944 (founding NTHL1 report): reports the same allele under the earlier nomenclature c.268C>T p.(Gln90*) (NM_002528.6), a homozygous germline nonsense mutation in seven affected individuals from three unrelated families with recessive inheritance of adenomatous polyposis and CRC; the allele was found only heterozygously in controls (MAF 0.0036, n=2,329) and segregated with a recessively inherited trait. The equivalence c.268C>T p.(Gln90*) (NM_002528.6) = c.244C>T p.(Gln82*) (NM_002528.7), ClinVar ID 192319, is stated explicitly in PMID:37727376.
• ClinVar (VCV000192319, variation 192319): the majority of submitters classify NM_002528.7(NTHL1):c.244C>T (p.Gln82Ter) as Pathogenic/Likely pathogenic; submission SCV001551973 states the NTHL1 Gln90Ter allele was identified 'in the compound heterozygous state with a canonical splice variant, confirmed to be in trans through familial testing' in an individual with colon cancer and multiple colon polyps.
• Generic ACMG/AMP rule applied verbatim because no gene-specific specification exists: PM3 (Pathogenic, Moderate) = 'For recessive disorders, detected in trans with a pathogenic variant' (Richards et al. 2015, PMID:25741868).

Still missing

• Trans partner in the best-characterized case (PMID:37727376 patient #5) is a novel in-frame deletion classified likely pathogenic rather than pathogenic, and the other compound-heterozygous partner (c.503T>C p.(Ile168Thr)) is a VUS; PM3 here rests on biallelic occurrence (trans or homozygous) rather than exclusively on a definitively pathogenic trans allele.
• No VCEP/ClinGen CSPEC specification exists for NTHL1, so the generic ACMG/AMP PM3 definition was applied at its default Moderate strength.

Sources

Reasoning

PP1 (co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease) is met at supporting strength. PMID:37727376 explicitly establishes that the NTHL1 variant originally reported by Weren et al. 2015 as c.268C>T p.(Gln90Ter) under the legacy NM_002528.6 transcript is the same allele as this case's NM_002528.7:c.244C>T p.(Gln82Ter) (ClinVar 192319), so the segregation data in PMID:25938944 apply to this variant. In that cohort the homozygous truncating allele was found in 7 clinically affected individuals from 3 unrelated families; Sanger sequencing showed three further affected individuals in families B and C were also homozygous, a heterozygous carrier in family B was not clinically affected, and all pedigrees were consistent with recessive inheritance of adenomatous polyposis, with the allele seen in controls only in the heterozygous state. Co-segregation with disease across multiple affected family members in a gene definitively known to cause NTHL1-associated polyposis satisfies PP1. Strength is kept at the ACMG/AMP 2015 default supporting: no governed, verified threshold for upgrading PP1 by number of meioses is provided in this case's documents, and because the disorder is autosomal recessive, homozygosity in affected relatives of carrier parents is partly Mendelian expectation and gives no basis to upgrade.

Rule

ACMG/AMP 2015 PP1 (Supporting): Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease (PMID:25741868).

Evidence considered

• PMID:37727376 states that the variant Weren et al. reported as c.268C>T p.(Gln90Ter) under NM_002528.6 is the same variant as NM_002528.7:c.244C>T p.(Gln82Ter), ClinVar ID 192319, which is exactly this case's allele.
• PMID:25938944 reports the homozygous NTHL1 p.Gln90Ter allele in seven affected individuals from three unrelated families; three additional clinically affected individuals in families B and C were homozygous on Sanger sequencing, a heterozygous carrier in family B was not clinically affected, and all pedigrees were in agreement with a recessively inherited trait with the allele found only heterozygously in controls.
• NTHL1 is definitively known to cause adenomatous polyposis/CRC: the homozygous loss-of-function germline state is established as causative, so the 'gene definitively known to cause the disease' requirement of PP1 is satisfied.
• No in-case proband, pedigree, or meioses exist in this case; PP1 is based solely on the three published recessive families.
• PP1 is retained at supporting strength (ACMG/AMP 2015 default); no publication supplying a governed upgrade threshold was available, and the autosomal recessive setting provides no basis to upgrade.

Still missing

• Segregation rests on three published families (PMID:25938944); no in-case pedigree, meioses, or family genotypes are available for this variant.
• No governed upgrade threshold for PP1 by number of meioses is available, so supporting strength is retained.

Sources

Reasoning

PS2 requires a confirmed de novo occurrence (both maternity and paternity confirmed) in a patient with the disease and no family history. No proband, trio, or parental genotype data exist anywhere in this case; the case is variant-level only (review_manifest.json and case_summary.json contain no patient, pedigree, or parental testing record). Beyond the absence of data, the substantive evidence points the other way: NTHL1-associated polyposis is an autosomal recessive disorder and the same truncating allele is documented as homozygous in clinically affected members of recessive families whose heterozygous carrier parents transmitted it (PMID:25938944), which is inherited rather than de novo. No source reports a de novo occurrence of c.244C>T p.(Gln82Ter). PS2 is therefore not satisfied.

Rule

ACMG/AMP 2015 PS2 (Strong): De novo (both maternity and paternity confirmed) in a patient with the disease and no family history (PMID:25741868).

Evidence considered

• No proband, trio, or parental testing record is present in the case files; the evidence bundle and case summary for this case carry variant-level data only (gene, HGVS, population frequency, predictors, ClinVar, literature).
• PMID:25938944 describes the same truncating allele as homozygous in 7 clinically affected individuals from 3 unrelated families, with heterozygous carrier parents and one unaffected heterozygous carrier, i.e. inherited recessively rather than arising de novo.
• PMID:37727376 establishes that the allele reported by Weren et al. as c.268C>T p.(Gln90Ter) (NM_002528.6) is this case's NM_002528.7:c.244C>T p.(Gln82Ter) (ClinVar 192319), so the recessive transmission described there applies to this variant.
• ClinVar submissions for variation 192319 record germline origin; no submitter reports a de novo occurrence.

Still missing

• No proband, trio, or parental genotype data are available in this case; a confirmed de novo occurrence of c.244C>T has not been reported by any source.

Sources

Reasoning

Not met. With no NTHL1-specific VCEP/CSPEC specification available, PM2 is assessed against the generic SVI supporting threshold (absent/rare, FAF <= 0.0001) using the all-comers gnomAD v2.1 / v4.1 totals. The variant is present at gnomAD v2.1 total AF 0.00141732 (400/282222 alleles) and gnomAD v4.1 total AF 0.00159187 (2568/1613200 alleles), roughly 14-16-fold above the 0.0001 threshold. The grpmax FAF, which is the value recommended for PM2 because it controls for founder/ancestry artifacts, is 0.00177304 (v2.1) and 0.00170283 (v4.1), also well above threshold. No population supports rarity: the variant is seen at 0.00354751 (89/25088) in v2.1 European (Finnish) and 0.00462552 (292/63128) in v4.1 European (Finnish), and it is present in multiple continental groups (e.g., African/African American AF 0.00032, Admixed American AF 0.00071). The disease-relevant non-cancer subsets are similarly too common (gnomAD v2.1 non-cancer exomes 0.00143101, 339/236896; gnomAD v3.1 non-cancer genomes 0.00135247, 200/147878). Because all relevant frequencies exceed 0.0001 by more than an order of magnitude, PM2 is not met.

Rule

PM2 (supporting): gnomAD all-comers FAF <= 0.0001 (SVI 2020, PMID:25741868); all observed frequencies exceed this.

Evidence considered

• gnomAD v2.1 all-comers total AF 0.00141732 (400/282222 alleles) and v4.1 all-comers total AF 0.00159187 (2568/1613200 alleles); both about 14-16-fold above the 0.0001 PM2 threshold.
• gnomAD v2.1 grpmax FAF 0.00177304 and v4.1 grpmax FAF 0.00170283; highest subpopulation European (Finnish) AF 0.00354751 (v2.1) and 0.00462552 (v4.1).
• gnomAD v2.1 continental subpopulations showing the variant in multiple ancestries: African/African American AF 0.00032 (8/24924), Admixed American AF 0.00071 (25/35432).
• gnomAD v2.1 non-cancer exomes AF 0.00143101 (339/236896) and gnomAD v3.1 non-cancer genomes AF 0.00135247 (200/147878) confirm the variant is not rare even in the non-cancer subsets.
• PM2 threshold used: FAF <= 0.0001, SVI recommendation to apply PM2 at supporting strength (PMID:25741868). No NTHL1 VCEP-specific threshold exists.
• PM2 is inapplicable here on frequency grounds regardless of the recessive inheritance of NTHL1, because the variant is a common recurrent founder allele rather than a rare singleton.

Sources

Reasoning

PM6 requires an assumed de novo occurrence without confirmation of maternity and paternity. No proband-level data are present in this case and no source reports c.244C>T p.(Gln82Ter) arising de novo. The only family-level information available shows the opposite direction of transmission: the equivalent legacy allele (c.268C>T p.Gln90Ter) was inherited from heterozygous carrier parents in autosomal recessive NTHL1 polyposis families (PMID:25938944, PMID:37727376). PM6 is therefore not satisfied.

Rule

ACMG/AMP 2015 PM6 (Moderate): Assumed de novo, but without confirmation of paternity and maternity (PMID:25741868).

Evidence considered

• No assumed de novo occurrence of c.244C>T is reported by any source consulted, and the case contains no proband-level record.
• PMID:25938944 shows the same allele transmitted from heterozygous carrier parents in recessive families, which is incompatible with an assumed de novo event.
• PMID:37727376 confirms the legacy c.268C>T p.(Gln90Ter) patients are this variant, so the recessive, parentally transmitted pattern applies here.

Still missing

• No proband-level or parental data are available; no source describes c.244C>T as an assumed de novo event.

Sources

Reasoning

Not met. PP5 may be applied only where a reputable source - here, an exact-variant ClinVar expert-panel (ClinGen VCEP) classification - supports a Pathogenic or Likely pathogenic call. The exact-variant ClinVar record for this allele is VCV000192319 (variation ID 192319, NM_002528.7(NTHL1):c.244C>T p.(Gln82Ter)). Its overall review status is 'criteria provided, single submitter' (ClinVar API also returns 'criteria provided, conflicting classifications', 1 star, 43-45 submissions), and ClinVar reports zero expert-panel submissions for this variant (case_summary submission_audit_summary expert_panel_submissions = 0; evidence.json source_data expert_panels = {}). No ClinGen NTHL1 VCEP exists at all (no CSPEC, gene_vcep_dir = null). Ordinary laboratory submissions (Baylor Genetics, Ambry, Invitae/Labcorp, Mayo, Quest, Sema4, Variantyx, etc.) and aggregate ClinVar labels are explicitly not eligible to trigger PP5, so the criterion's trigger condition is absent. PP5 is therefore not met.

Rule

PP5 (supporting, per governing instruction): only an exact-variant ClinVar expert-panel (ClinGen VCEP) Pathogenic/Likely pathogenic classification triggers PP5; ordinary laboratory submissions and aggregate ClinVar labels do not. No such expert-panel classification exists here.

Evidence considered

• ClinVar VCV000192319 (variation ID 192319), NM_002528.7(NTHL1):c.244C>T (p.Gln82Ter): overall review status 'criteria provided, single submitter' per the scraped record; ClinVar API returns 'criteria provided, conflicting classifications' with review_stars = 1 and 43 submissions / 43 submitters, last updated 2026-06-20.
• Expert-panel check: ClinVar source_data 'expert_panels' is empty ({}), and the case submission audit reports expert_panel_submissions = 0 out of 45 submissions. No ClinGen expert panel (VCEP) classification of this exact variant exists.
• Submission composition (aggregate, not expert): 36 clinical laboratories reported Pathogenic, 2 Likely pathogenic, 1 pathogenic, 1 Uncertain significance; submitters are clinical/diagnostic laboratories (OMIM, Baylor Genetics, PreventionGenetics, Ambry Genetics, Invitae/Labcorp, Mayo Clinic Laboratories, Quest Diagnostics, Sema4, Variantyx, Dasa, SA Pathology, Rigshospitalet, ADN Uruguay, Hospital Israelita Albert Einstein, Ruhr University Bochum, TU Dresden, UMC Groningen, Sinai Health System, Mendelics). Ordinary laboratory submissions and aggregate ClinVar labels do not trigger PP5.
• No NTHL1 ClinGen CSPEC or VCEP specification exists for this gene (vcep_materials: 'no cspec url available', gene_vcep_dir = null), so there is no expert-panel assertion for this variant at any strength.

Sources

Reasoning

Not met. There is no NTHL1-specific VCEP/CSPEC or gene-specific population-frequency specification in this case (framework_mode = generic_acmg, cspec not found), so the generic ACMG/AMP 2015 stand-alone benign threshold (BA1, AF >= 0.05) applies and the all-comers gnomAD v2.1 / v4.1 totals are the default source. gnomAD v2.1 all-comers: total AF 0.00141732 (400/282222 alleles, 0 homozygotes); exome AF 0.00141113 (354/250862); genome AF 0.00146684 (46/31360). gnomAD v4.1 all-comers: total AF 0.00159187 (2568/1613200 alleles, 1 homozygote); exome AF 0.00161813 (2364/1460944); genome AF 0.00133985 (204/152256). The highest population frequency observed is European (Finnish): 0.00354751 (89/25088) in v2.1 and 0.00462552 (292/63128) in v4.1; grpmax FAF is 0.00177304 (v2.1) and 0.00170283 (v4.1). Every one of these values is more than an order of magnitude below the 0.05 stand-alone threshold, so the variant is far too rare for a stand-alone benign classification. The non-cancer subsets are lower still (gnomAD v2.1 non-cancer exomes AF 0.00143101, 339/236896; gnomAD v3.1 non-cancer genomes AF 0.00135247, 200/147878) and would not change the conclusion even if a VCEP had mandated them. The variant is a known recurrent, founder-enriched NTHL1 truncating allele, which explains why its frequency is not vanishingly rare, but recurrence in this range does not approach a stand-alone benign frequency.

Rule

BA1: gnomAD all-comers AF >= 0.05 (generic ACMG/AMP 2015, PMID:25741868); no NTHL1 VCEP override.

Evidence considered

• gnomAD v2.1 all-comers (16-2096239-G-A): total AF 0.00141732 (400/282222 alleles, 0 homozygotes); exome AF 0.00141113 (354/250862, 0 hom); genome AF 0.00146684 (46/31360, 0 hom); highest subpopulation European (Finnish) AF 0.00354751 (89/25088, 0 hom); grpmax FAF 0.00177304.
• gnomAD v4.1 all-comers (16-2046238-G-A): total AF 0.00159187 (2568/1613200 alleles, 1 homozygote); exome AF 0.00161813 (2364/1460944, 1 hom); genome AF 0.00133985 (204/152256, 0 hom); highest subpopulation European (Finnish) AF 0.00462552 (292/63128, 0 hom); grpmax FAF 0.00170283.
• gnomAD v2.1 non-cancer exomes: AF 0.00143101 (339/236896 alleles, 0 homozygotes).
• gnomAD v3.1 non-cancer genomes: AF 0.00135247 (200/147878 alleles, 0 homozygotes); grpmax FAF 0.00156202.
• BA1 stand-alone threshold used: AF >= 0.05 (ACMG/AMP 2015 sequence variant interpretation standards, PMID:25741868). No observed frequency reaches this value.
• No NTHL1 VCEP or ClinGen CSPEC specification exists for this gene (cspec found = false; gene_vcep_dir = null), so the all-comers gnomAD v2.1/v4.1 totals are the default population source and the non-cancer subsets were reviewed only for completeness.
• PMID:37727376 establishes c.244C>T (p.Gln82Ter) as the recurrent founder NTHL1 truncating allele (originally c.268C>T under NM_002528.6, ClinVar ID 192319), explaining the non-trivial Northern European frequency; recurrence at this level is not a stand-alone benign criterion.

Sources

Reasoning

Not met. With no NTHL1-specific VCEP/CSPEC specification available, the generic strong benign threshold (BS1, AF >= 0.01; generic SVI default, per-gene derivation preferred per Whiffin et al. 2017, PMID:28518168) applies to the all-comers gnomAD v2.1 / v4.1 totals. gnomAD v2.1 total AF 0.00141732 (400/282222) and gnomAD v4.1 total AF 0.00159187 (2568/1613200); grpmax FAF 0.00177304 (v2.1) and 0.00170283 (v4.1). The variant is enriched in European (Finnish) individuals at 0.00354751 (v2.1) and 0.00462552 (v4.1), consistent with its reported Dutch founder origin, but even this strongest ancestry-specific frequency is about half the 0.01 threshold and the all-comers count is roughly 6-fold lower. The disease-relevant non-cancer subsets are lower still (gnomAD v2.1 non-cancer exomes 0.00143101; gnomAD v3.1 non-cancer genomes 0.00135247) and do not reach the threshold. No population, including the Finnish subpopulation, reaches AF >= 0.01, so BS1 is not met.

Rule

BS1: gnomAD all-comers AF >= 0.01 (generic SVI default, PMID:28518168); no NTHL1 VCEP override.

Evidence considered

• gnomAD v2.1 all-comers total AF 0.00141732 (400/282222 alleles, 0 homozygotes); highest subpopulation European (Finnish) AF 0.00354751 (89/25088); grpmax FAF 0.00177304.
• gnomAD v4.1 all-comers total AF 0.00159187 (2568/1613200 alleles, 1 homozygote); highest subpopulation European (Finnish) AF 0.00462552 (292/63128); grpmax FAF 0.00170283.
• gnomAD v2.1 non-cancer exomes AF 0.00143101 (339/236896) and gnomAD v3.1 non-cancer genomes AF 0.00135247 (200/147878) were reviewed as the disease-relevant non-cancer subsets; neither reaches 0.01 and neither was mandated by an NTHL1 VCEP (none exists).
• BS1 threshold used: AF >= 0.01, generic SVI default with per-gene derivation preferred (Whiffin et al. 2017, PMID:28518168). No NTHL1-specific derivation is available, so the generic 0.01 value was applied.
• PMID:37727376 documents the founder/recurrent nature of c.244C>T (p.Gln82Ter), which accounts for the elevated Northern European frequency but does not raise it to the strong benign threshold.

Sources

Reasoning

Not met. BS2 requires the variant to be observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked disorder with full penetrance expected at an early age, i.e. a homozygote in a population cohort who has escaped a severe, early-onset, fully penetrant disease. Homozygote counts for this variant are 0 in gnomAD v2.1 (total and every subpopulation) and 1 in gnomAD v4.1 exomes (the single homozygote falls in the NFE_XX subset; total homozygotes = 1). NTHL1 tumor syndrome is an autosomal recessive, adult-onset cancer predisposition condition with incomplete penetrance: PMID:37727376 reports homozygous carriers presenting with colorectal polyposis and colorectal cancer at ages 47-63, i.e. adulthood, and notes that heterozygous carriers are not at clearly increased risk, while homozygous carriers show a broad and variable tumor spectrum (colorectal, breast, endometrial, pancreatic, basal cell). A single homozygote in an adult population database is therefore compatible with the known penetrance profile of this adult-onset recessive condition and does not meet the BS2 requirement that the disorder be fully penetrant and expected at an early age. gnomAD also does not report the phenotype or health status of that homozygote, so it cannot be asserted to be a 'healthy adult' for a fully penetrant early-onset disorder. BS2 is therefore not met.

Rule

BS2: observed homozygous in a healthy adult for a recessive disorder with full penetrance expected at an early age (generic ACMG/AMP 2015, PMID:25741868); criterion conditions not satisfied.

Evidence considered

• gnomAD v2.1 (all-comers) homozygote count = 0 (total, exome and genome); every v2.1 subpopulation homozygote count = 0.
• gnomAD v4.1 (all-comers) homozygote count = 1, occurring in the exome NFE_XX subset (611624 alleles, 1 homozygote; total v4.1 homozygotes = 1); genome homozygote count = 0.
• gnomAD-Canada v1.0 (HostSeq genomes) is not available for this variant.
• BS2 genetic mechanism requirement: observed in a healthy adult for a recessive/dominant/X-linked disorder with full penetrance expected at an early age (ACMG/AMP 2015, PMID:25741868).
• PMID:37727376: homozygous c.244C>T carriers presented with colorectal polyposis and colorectal cancer at ages 47-63 and a variable multi-tumor spectrum (colorectal, breast, endometrial, pancreatic, basal cell carcinoma); heterozygous carriers were not at clearly increased risk. This defines NTHL1 tumor syndrome as an adult-onset, incompletely penetrant recessive condition, so a single population homozygote does not satisfy BS2.
• The gnomAD record does not report phenotype or health status for the single homozygote, so 'healthy adult with a fully penetrant early-onset disease' cannot be substantiated.

Sources

Reasoning

BS3 is not met. BS3 requires well-established functional studies showing no damaging effect on protein function; no such assay exists for this variant. The only variant-specific functional assay available is the RNA expression / nonsense-mediated-decay assay in carrier-derived B lymphocytes (PMID:25938944), and it demonstrates the opposite of a benign effect: loss of the NTHL1 transcript via NMD, i.e. a deleterious functional consequence, corroborated in vivo by the base-excision-repair-deficient COSMIC SBS30 mutational signature in tumours of homozygous carriers (PMID:37727376). PMID:31285513 reported a monoallelic c.268C>T (legacy name for this variant) carrier with normal/elevated wild-type transcript, but this reflects the absence of NTHL1 haploinsufficiency in the heterozygous state and does not demonstrate that the variant protein retains function; biallelic NTHL1 loss is the disease mechanism, so a monoallelic RNA result is not BS3-supporting. No assay measured NTHL1 glycosylase activity directly. A clinical submission (SCV005902063) notes no well-established functional studies for the variant, which is a functional-evidence gap rather than BS3 evidence. Accordingly BS3 is not met.

Rule

ACMG/AMP 2015 BS3 (PMID:25741868): well-established functional studies show no damaging effect on protein function or splicing. Generic fallback (no NTHL1 VCEP/CSPEC). Not met: the only variant-specific functional data show loss of transcript via nonsense-mediated decay, i.e. a damaging effect.

Evidence considered

• PMID:25938944 (Weren et al. 2015): the variant-specific B-lymphocyte qRT-PCR/NMD assay shows the nonsense allele 'results in nonsense-mediated decay', i.e. a damaging effect on the gene product, which is the inverse of the benign functional effect BS3 requires.
• PMID:37727376: tumours from homozygous carriers of this exact variant carry the COSMIC SBS30 mutational signature and a C>T transition bias, an in-vivo loss-of-function phenotype of defective base-excision repair, further arguing against a neutral/benign functional effect.
• PMID:31285513: examined a monoallelic c.268C>T p.(Gln90*) (legacy name for this variant) carrier and reported normal/elevated wild-type transcript levels, concluding there was no NTHL1 haploinsufficiency; this NEGATIVE RNA result concerns the heterozygous state and does not show the variant protein is functional, so it does not support BS3.
• ClinVar submission SCV005902063 (Catalan Institute of Oncology) states 'to our knowledge, no well-established functional studies have been reported for this variant'; this is an absence-of-evidence statement, not functional evidence of a benign effect, so BS3 is not met rather than not_assessed.
• Strength/calibration: no NTHL1 VCEP/CSPEC specification exists, so generic ACMG/AMP 2015 governs; no benign-direction functional evidence of any strength was identified.

Still missing

• No validated functional assay demonstrating normal NTHL1 function for this allele exists, and none is expected given the premature stop codon and the demonstrated NMD; BS3 remains unsupported rather than merely unassessed.

Sources

Reasoning

BS4 requires lack of segregation of the variant with disease in affected members of a family. No non-segregation data exist for this variant, and the only family-level evidence available points in the opposite direction: the equivalent legacy allele (c.268C>T p.Gln90Ter) was homozygous in affected members of three recessive NTHL1 polyposis families and heterozygous in an unaffected carrier (PMID:25938944, with the transcript equivalence established by PMID:37727376). No affected relative was reported to lack the variant, and no source describes a non-segregating pedigree. BS4 is therefore not met.

Rule

ACMG/AMP 2015 BS4 (Strong): Lack of segregation in affected members of a family (PMID:25741868).

Evidence considered

• PMID:25938944 reports homozygous carriers among affected family members (7 affected across 3 families, plus 3 further affected homozygotes in families B and C) and an unaffected heterozygous carrier in family B, which is segregation, not non-segregation.
• PMID:37727376 confirms that the c.268C>T p.(Gln90Ter) families described by Weren et al. carry this case's c.244C>T p.(Gln82Ter) allele, so the segregation pattern applies to this variant.
• No source reports an affected individual carrying the phenotype but lacking this variant, and no non-segregating pedigree is available; the criterion has no supporting evidence.

Still missing

• No pedigree with genotyped affected and unaffected relatives is available in this case; BS4 cannot be assessed on in-case meioses, and no non-segregation has been reported in the literature.

Sources

Reasoning

BP2 is met when a variant is observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder, or when it is observed in cis with a pathogenic variant in any inheritance pattern. NTHL1 tumor syndrome is autosomal recessive, not a fully penetrant dominant disorder, so the trans-with-pathogenic-in-dominant-disease branch does not apply. For the cis branch, the available phase data indicate the opposite configuration: PMID:37727376 places c.244C>T p.(Gln82Ter) in trans with c.435_446del p.(Leu146_Arg149del) (with family testing confirming that the two variants are on opposite alleles), and PMID:25938944 reports the same allele (as c.268C>T p.(Gln90*)) homozygously in three recessive-inheritance families. No source places c.244C>T in cis with a pathogenic variant, and biallelic (trans) occurrence is evidence in the pathogenic, not benign, direction. BP2 is therefore not met.

Rule

BP2 (Benign, Supporting): observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder, or observed in cis with a pathogenic variant in any inheritance pattern (Richards et al. 2015, PMID:25741868).

Evidence considered

• PMID:37727376 states that NTHL1 tumor syndrome 'is inherited in an autosomal recessive manner' and reports c.244C>T only in biallelic states (homozygous or compound heterozygous); it is therefore not in cis with a pathogenic variant, and the gene is not a fully penetrant dominant disease gene.
• PMID:37727376, Figure 1C / Results 3.1: family testing showed c.244C>T and c.435_446del p.(Leu146_Arg149del) are in trans (relatives carry only one of the two variants), which is the opposite of the cis configuration BP2 requires.
• PMID:25938944: the same allele (reported as c.268C>T p.Gln90*) was homozygous in seven affected individuals from three unrelated families showing recessive inheritance, again excluding a cis arrangement.
• No reviewed source places c.244C>T in cis with a pathogenic variant, and NTHL1-associated disease is recessive, so neither branch of BP2 is satisfied.
• Generic ACMG/AMP rule applied verbatim because no gene-specific specification exists: BP2 (Benign, Supporting) = observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder, or observed in cis with a pathogenic variant in any inheritance pattern (Richards et al. 2015, PMID:25741868).

Sources

Reasoning

Not met. Under the governing rule, BP6 may be applied only where an exact-variant ClinVar expert-panel classification supports a Benign or Likely benign call; ordinary laboratory submissions, aggregate ClinVar labels and non-expert assertions must never trigger it. ClinVar VCV000192319 (NM_002528.7(NTHL1):c.244C>T p.(Gln82Ter)) has zero expert-panel submissions (submission_audit_summary expert_panel_submissions = 0 out of 45; evidence.json expert_panels = {}), a review status of 'criteria provided, single submitter' (1 star; ClinVar API additionally shows conflicting classifications), and no Benign/Likely benign submission of any kind - the submission summary is 36 clinical laboratories Pathogenic, 2 Likely pathogenic, 1 pathogenic and 1 Uncertain significance. No ClinGen NTHL1 VCEP exists (no CSPEC, gene_vcep_dir = null), so there is no expert-panel benign assertion of this variant anywhere. The trigger condition for BP6 is therefore absent and the criterion is not met.

Rule

BP6 (supporting, per governing instruction): only an exact-variant ClinVar expert-panel (ClinGen VCEP) Benign/Likely benign classification triggers BP6; ordinary laboratory submissions, aggregate labels and non-expert assertions do not. No such classification exists here.

Evidence considered

• ClinVar VCV000192319 (variation ID 192319), NM_002528.7(NTHL1):c.244C>T (p.Gln82Ter): review status 'criteria provided, single submitter' (1 star); ClinVar API returns 'criteria provided, conflicting classifications'. Submission summary contains no Benign or Likely benign assertion (36 clinical laboratories Pathogenic, 2 Likely pathogenic, 1 pathogenic, 1 Uncertain significance).
• Expert-panel check: source_data 'expert_panels' is empty ({}), and the case submission audit reports expert_panel_submissions = 0 of 45 submissions; no ClinGen expert panel (VCEP) has classified this variant.
• No NTHL1 ClinGen CSPEC or VCEP specification exists (vcep_materials: 'no cspec url available'; gene_vcep_dir = null), so no expert-panel benign assertion for this variant is available to consult.

Sources

Reasoning

Not assessed. PP4 requires that the phenotype (or family history) of the individual being evaluated is highly specific for a disease with a single genetic etiology. This case was submitted as a variant only (NM_002528.7:c.244C>T); review_manifest.json, case_summary.json and prefetch.json contain no proband clinical description, no HPO terms, no family history and no tumour phenotype, so there is no patient-level phenotype against which PP4 can be judged. Published cohort data do show the biallelic form of this allele in individuals with the highly specific NTHL1 tumour syndrome phenotype (multiple colorectal adenomas/polyposis with a characteristic multi-tumour spectrum, including colorectal cancer and basal-cell carcinoma; PMID:37727376, PMID:25938944), but that is phenotype information about published cohorts, not about the case under classification, and PP4 may not be imported from gene- or cohort-level phenotype. Returned as not_assessed so that the missing proband phenotype is recorded rather than silently converted into a not_met verdict.

Rule

PP4 (generic ACMG/AMP 2015, PMID:25741868): patient's phenotype or family history highly specific for a disease with a single genetic etiology; not evaluable without proband phenotype data.

Evidence considered

• Case record review: review_manifest.json, case_summary.json (compact_evidence) and prefetch.json contain variant-level data only (HGVS, gene, ClinVar, gnomAD, SpliceAI, evidence signals); no proband phenotype, HPO code, family history, age of onset or tumour phenotype is recorded anywhere in the case.
• PP4 criterion requirement (generic ACMG/AMP 2015, PMID:25741868): the patient's phenotype or family history is highly specific for a disease with a single genetic etiology; this is a patient-level criterion.
• Published phenotype context (not usable as the case's own phenotype): PMID:37727376 describes homozygous carriers of this exact allele with colorectal polyposis (<10 to >100 polyps), colorectal cancer at ages 47-63, breast cancer, endometrial lesions, pancreatic cancer and basal-cell carcinoma; PMID:25938944 (original report of the same allele under the older name c.268C>T p.(Gln90Ter), NM_002528.6) describes 7 affected individuals from 3 families with multiple colonic adenomas. These establish that biallelic c.244C>T produces a recognisable NTHL1 tumour syndrome phenotype, which would be relevant to PP4 for an affected proband if one were provided.
• No NTHL1 ClinGen CSPEC or VCEP exists, so no gene-specific PP4 phenotype checklist is available to apply instead.

Still missing

• No proband clinical phenotype, HPO terms, family history, or tumour spectrum is available for this case; PP4 cannot be concluded without it.
• If the referring clinician supplies the indication for testing (e.g. multiple colorectal adenomas/polyposis, early-onset colorectal cancer, NTHL1 tumour syndrome), PP4 should be re-evaluated against the known NTHL1 phenotype spectrum.

Sources

Reasoning

Not assessed. BP5 requires that the variant be found in a case that has an alternative molecular basis for the disease, i.e. it needs a genotyped, phenotyped individual in whom a different genetic cause explains the presentation. This case contains no proband genotype/phenotype, no co-occurring variant information, and no molecular diagnosis, so there is no affected individual in which an alternate molecular basis could be demonstrated. The only near-relevant literature observation is PMID:31285513, in which the same allele under the older name (c.268C>T p.(Gln90Ter), NM_002528.6) was found in the monoallelic state in a double POLE/NTHL1 carrier whose NTHL1 involvement was discounted for lack of haploinsufficiency - that is an argument about monoallelic NTHL1 not being disease-causing per se in that individual, not an alternate-molecular-basis finding for the case being classified, and it is a single case. PMID:33193653 reports the same allele monoallelically (c.268C>T p.(Gln90*) in Family 136) with no stated alternate genetic cause. BP5 therefore cannot be concluded either way and is returned as not_assessed.

Rule

BP5 (generic ACMG/AMP 2015, PMID:25741868): variant found in a case with an alternate molecular basis for disease; not evaluable without case-level genotype/phenotype data.

Evidence considered

• Case record review: no proband genotype/phenotype, co-occurring variant or established molecular diagnosis is recorded in review_manifest.json, case_summary.json or the evidence bundle, so there is no individual in which an alternate molecular basis could be shown.
• BP5 criterion requirement (generic ACMG/AMP 2015, PMID:25741868): the variant is found in a case with an alternate molecular basis for disease (e.g. a different gene explaining the phenotype); this is a case-level criterion.
• PMID:31285513 (158-patient Spanish attenuated adenomatous polyposis cohort): the monoallelic truncating NTHL1 c.268C>T p.(Gln90*) - the same allele as c.244C>T p.(Gln82Ter) under NM_002528.7 - was investigated in a double POLE/NTHL1 carrier and its involvement discarded because of absent NTHL1 haploinsufficiency. This is a single monoallelic observation used by those authors to argue against NTHL1 haploinsufficiency, not evidence of an alternate molecular basis in the case under evaluation.
• PMID:33193653: monoallelic NTHL1 c.268C>T p.(Gln90*) reported in Family 136 (patient #136;12, sigmoid colon cancer at 49, basal cell carcinoma at 51, two adenomas) with no alternate genetic cause stated; does not satisfy the BP5 trigger for this case.
• No NTHL1 ClinGen CSPEC or VCEP exists, so no gene-specific BP5 rule is available to apply instead.

Still missing

• No proband-level data (co-occurring variants, molecular diagnosis, or genotype-phenotype attribution) are available for this case; BP5 requires a case in which a different genetic cause explains the disease.

Sources