| Source / year / DOI | Design and denominator | Exact key numerical observation | Main limitations |
|---|---|---|---|
| **Badar et al., 2023** — [10.3324/haematol.2023.282867](https://doi.org/10.3324/haematol.2023.282867) | Retrospective, clinically ascertained sequencing cohort: **4,524** patients evaluated for suspected/known myeloid neoplasms; **107** had DDX41 variants (**44 P/LP**, **63 VUS**, including 11 with both). (pqac-00000011) | DDX41 P/LP variants occurred in **44/4,524 (0.9%)** and VUS in **63/4,524 (1.4%)**. Among high-risk MDS/AML, median overall survival was **63.4 months** with P/LP versus **55.7 months** with VUS (*P*=0.93). (pqac-00000011, pqac-00000063) | Referral/testing cohort, not population-based; many variants were presumed germline from variant allele fraction rather than confirmed in nonhematopoietic tissue; VUS inclusion complicates causal inference. |
| **Kusne et al., 2025** — [10.1111/bjh.20018](https://doi.org/10.1111/bjh.20018) | Retrospective Mayo Clinic DDX41 germline-predisposition cohort: **195** individuals; **42.3% P/LP** and **57.6% VUS**. (pqac-00000010) | Within this selected cohort, diagnoses included **MDS 79/195 (40.5%)**, **AML 40/195 (20.5%)**, and **MPN 13/195 (6.7%)**; these are cohort case proportions, not population prevalence. Among myeloid-neoplasm patients, **78.8%** had a normal karyotype. (pqac-00000010, pqac-00000035) | Clinic-based ascertainment enriched for disease; most participants were White (**92%**); more than half carried VUS; retrospective categories overlap with other malignancy histories and do not establish longitudinal transition rates. |
| **Tesi et al., 2025** — [10.1158/1078-0432.CCR-24-4251](https://doi.org/10.1158/1078-0432.CCR-24-4251) | Prospective multicenter Swedish validation of Nordic germline-testing criteria: **85** selected patients with AML, MDS, thrombocytopenia, or related phenotypes. | P/LP germline variants were found in **30/85 (35%)**; **DDX41 accounted for 13/30 (43%)** positive cases. Yield was **6% (1/16)** under family-history criteria versus **52% (17/33)** under tumor cytogenetic/molecular criteria. (pqac-00000075) | Only patients already meeting investigation criteria were enrolled; therefore, the 35% yield is not prevalence among all AML/MDS patients. Cohort size and national testing practices may limit generalizability. |
| **Hendricks et al., 2025**; online 6 Nov 2024 — [10.1038/s41375-024-02436-y](https://doi.org/10.1038/s41375-024-02436-y) | Genome-first population analysis; UK Biobank included **469,595** exomes and **815** DDX41 P/LP heterozygotes. | DDX41 P/LP-carrier frequency was **815/469,595**, approximately **1 in 576**. Hematologic malignancy was recorded in **43/815 (5.2%; 1 in 19)** carriers; myeloid-malignancy risk was elevated (**OR 5.7**, 95% CI 3.9–8.3). (pqac-00000014, pqac-00000016) | Electronic-record and tumor-registry outcomes may miss diagnoses; UK Biobank has healthy-volunteer and ancestry biases; 5.2% is observed record-based penetrance over available follow-up, not lifetime penetrance. |
| **Korotev et al., 2025** — [10.3324/haematol.2024.286887](https://doi.org/10.3324/haematol.2024.286887) | Family-based cohort: **93 P/LP DDX41 carriers from 43 families**; marrow was reviewed in 29 carriers. | Estimated hematologic-malignancy penetrance was **50/93 (54%) by age 90**. Among reviewed marrows, small hypolobated megakaryocytes and macrocytic erythrocytes were reported in **75%** and **50%**, respectively. (pqac-00000064, pqac-00000067) | Families were clinically selected for suspected predisposition, which can inflate penetrance; only a subset had marrow review; baseline dysplasia can cause MDS overdiagnosis and requires clinical context. |
| **Stepanchick et al., 2024** — [10.1038/s41375-024-02304-9](https://doi.org/10.1038/s41375-024-02304-9) | Primary experimental study using hematopoietic-specific **Ddx41+/−** mice, competitive bone-marrow transplantation, transcriptomics, and combined **Ddx41+/−;Trp53+/−** mice. | Ddx41 haploinsufficiency caused only mild persistent red-cell reduction at steady state, but transplantation stress exposed HSPC fitness defects and age-related hematopoietic failure; reduced *Trp53* dosage rescued fitness yet promoted MDS/AML-like malignancies. (pqac-00000020, pqac-00000024) | Mouse transplantation stress is not equivalent to natural human aging; heterozygous Ddx41 loss alone did not reproduce overt human malignancy; the cooperating *Trp53* model represents only one possible secondary pathway. |
| **Bi et al., 2025** — [10.1038/s41467-025-62307-7](https://doi.org/10.1038/s41467-025-62307-7) | Mechanistic study spanning conditional mouse knockouts, a DDX41-mutated human sample, and human iPSC-derived bone-marrow organoids. | DDX41 deficiency increased G-quadruplex accumulation, erythroid genome instability, ribosomal defects, and p53 signaling; **cGas deficiency rescued embryonic lethality** in hematopoietic-specific *Ddx41*-knockout mice. (pqac-00000039) | Primarily preclinical; complete knockout is more severe than typical human heterozygous germline disease; rescue of mouse lethality does not establish that cGAS inhibition prevents human MDS/AML. |


*Table: Primary clinical and experimental evidence is organized by design, denominator, exact quantitative finding, and major limitation. Cohort proportions are explicitly distinguished from population frequency and lifetime penetrance.*