| Domain | Key quantitative finding | Evidence type/source (author/year) | DOI/URL |
|---|---|---|---|
| Epidemiology and anemia | 232 patients across 24 centers; cold versus warmer climates: prevalence 20 versus 5 per million and incidence 1.9 versus 0.48 per million/year. Mean baseline Hb 9.3 g/dL; 27% had Hb <8 g/dL. | Multinational observational cohort; Berentsen et al., 2020 (pqac-00000016) | [10.1182/blood.2020005674](https://doi.org/10.1182/blood.2020005674) |
| Clinical phenotypes | Anemia affects ~90%; cold-induced circulatory symptoms occur in 57–91%; acute hemolytic exacerbations in ~74%; fatigue was reported by 90% in a selected patient survey. | Clinical synthesis; Berentsen et al., 2024 (pqac-00000011) | [10.1080/17474086.2024.2372333](https://doi.org/10.1080/17474086.2024.2372333) |
| Thrombosis and mortality | At 1 year, thromboembolism occurred in 7.2% of CAD patients versus 1.9% of controls; at 5 years, 11.5% versus 7.8%. Median survival was 8.5 years; adjusted mortality HR 1.84 (95% CI 1.10–3.06). | Danish national-registry cohort; Bylsma et al., 2019 (pqac-00000017) | [10.1182/bloodadvances.2019000476](https://doi.org/10.1182/bloodadvances.2019000476) |
| Somatic genetics | In 18 patients, nonsynonymous mutations occurred in **KMT2D** 12/18 (67%), **CARD11** 6/18 (33%), and **CXCR4** 4/18 (22%); CARD11/CXCR4 lesions correlated with lower Hb. | Whole-exome/targeted genomic study; Małecka et al., 2021 (pqac-00000028) | [10.1002/ajh.26205](https://doi.org/10.1002/ajh.26205) |
| Transcriptomics | RNA-seq of clonal B cells from 12 patients versus IgM-memory B cells from 4 controls identified 93 differentially expressed genes; **CR1/CD35** was reduced ~11-fold, with protein reduction confirmed by flow cytometry. | Human RNA-seq and flow-cytometry study; Małecka et al., 2024 (pqac-00000024) | [10.1093/cei/uxad135](https://doi.org/10.1093/cei/uxad135) |
| Diagnostic criteria | Confirmed CAD required chronic hemolysis, monospecific C3d-positive DAT, cold-agglutinin titer >64, and exclusion of overt lymphoma, active cancer, and recent *Mycoplasma pneumoniae* or EBV infection. | Cohort case-definition criteria; Berentsen et al., 2020 (pqac-00000022) | [10.1182/blood.2020005674](https://doi.org/10.1182/blood.2020005674) |
| Rituximab–bendamustine | Responses occurred in 35/45 (78%), including complete responses in 24/45 (53%); median response duration was not reached after 88 months, and estimated 5-year sustained remission was 77%. | Long-term observational treatment follow-up; Berentsen et al., 2020 (pqac-00000015, pqac-00000016) | [10.1182/blood.2020005674](https://doi.org/10.1182/blood.2020005674) |
| CARDINAL | Sutimlimab primary endpoint achieved by 13/24 (54%; 95% CI 33–74); mean Hb rose 1.2 g/dL at week 1 and 2.3 g/dL by week 3; bilirubin normalized by week 3. | Prospective phase 3 single-arm trial; summarized by Weitz, 2023 (pqac-00000034) | [10.17925/OHR.2023.19.1.35](https://doi.org/10.17925/OHR.2023.19.1.35) |
| CADENZA | Composite response occurred in 16/22 (72.7%) with sutimlimab versus 3/20 (15%) with placebo; median Hb increase was 2.7 g/dL. | Randomized placebo-controlled phase 3 trial; summarized by Costa et al., 2025 (pqac-00000032) | [10.3389/fimmu.2025.1624667](https://doi.org/10.3389/fimmu.2025.1624667) |
| Long-term sutimlimab | Among 39 CADENZA Part B patients, 32 completed treatment; median exposure was 99 weeks. Mean on-treatment Hb remained ≥11.0 g/dL versus 9.3 at baseline, and bilirubin ≤20.0 μmol/L versus 35.0 at baseline; hemolysis recurred after withdrawal. | Phase 3 open-label extension; Röth et al., published August 2024 (pqac-00000001) | [10.1016/j.eclinm.2024.102733](https://doi.org/10.1016/j.eclinm.2024.102733) |


*Table: Compact evidence table covering epidemiology, clinical burden, molecular findings, diagnostic criteria, and major treatment outcomes in cold agglutinin disease. It emphasizes cohort and trial-level quantitative findings without supplying unverified PMIDs.*