| Domain | Best current finding | Evidence type/strength | Key source/date |
|---|---|---|---|
| Disease/gene & inheritance | RAD21-related Cornelia de Lange syndrome is an ultra-rare cohesinopathy with an attenuated but variable CdLS phenotype; inheritance is typically autosomal dominant, with both de novo and familial heterozygous variants reported, plus rare biallelic RAD21 disease outside classic CDLS4 framing | Human cohort + landmark discovery; strong for causality | Deardorff et al., *Am J Hum Genet* 2012; Krab et al., *Hum Genet* 2020 (pqac-00000001, pqac-00000028) |
| Phenotype cohort | Best delineation includes 49 individuals from 33 families with RAD21 alterations; 29 had full clinical information | Human multicenter cohort; strongest phenotype source | Krab et al., *Hum Genet* 2020 (pqac-00000028) |
| Key phenotype frequencies | Among fully phenotyped RAD21 cases: 55% had normal or mildly impaired cognition; affected behavioral domains in 56%; anxiety 53%; ASD-like features 35%; ADHD/hyperactivity 35%; aggression and self-injury each 6% where assessed | Human cohort; moderate-strong, but behavior often based on clinical report rather than formal testing | Krab et al., *Hum Genet* 2020 (pqac-00000024, pqac-00000025, pqac-00000027) |
| 2023 variant reports | Novel de novo RAD21 c.1722_1723delTG (p.Gly575SerfsTer2) caused classic-scoring CDLS4 with PDA/VSD, cleft palate, laryngomalacia, prenatal growth retardation; separate 13-year-old girl had de novo heterozygous intragenic deletion of RAD21 exons 9-14 detected by high-resolution CMA, without reported neurodevelopmental disorder at diagnosis | Human case reports; useful expansion of spectrum, limited n | De Falco et al., *Genes* 2023; Abarca-Barriga et al., *Genes* 2023 (pqac-00000031, pqac-00000029, pqac-00000030) |
| Diagnostic yield | In 716 CdLS probands, 422/716 (~59%) received a molecular diagnosis; RAD21 accounted for 6/422 solved cases (~1%) | Large human genomic cohort; strong for relative contribution | Kaur et al., *Am J Med Genet A* 2023 (pqac-00000046) |
| Mechanism | Pathogenic RAD21 variants alter cohesin interfaces, perturb chromatin organization/transcription, impair DNA-damage response/repair, and in models disrupt zygotic genome activation, neural crest migration, and mesoderm/NMP development | Mixed human cell + zebrafish functional evidence; moderate-strong but partly inferred for human developmental pathology | Deardorff et al. 2012; Meier et al. 2018; Schuster et al. 2015; Labudina et al. 2024 (pqac-00000002, pqac-00000004, pqac-00000010, pqac-00000039, pqac-00000012) |
| Diagnosis | CdLS diagnosis remains clinical first, using consensus scoring; RAD21 cases often fall in non-classic/milder range. Molecular workup should include multigene sequencing plus CNV analysis; if negative and suspicion remains, test for mosaicism in non-blood tissue and consider broader sequencing | International consensus + cohort data; strong for practice | Kline et al. 2018 consensus; Kaur et al. 2023; Gruca-Stryjak et al. 2024 (pqac-00000037, pqac-00000036, pqac-00000046, pqac-00000006) |
| Treatment/trials | No approved RAD21-specific or disease-modifying therapy. Care is multidisciplinary/supportive. Current CdLS trials target broader CdLS populations and are not RAD21-specific; examples include lithium, behavioral intervention, NAC, and autonomic assessment studies | Consensus care + registry/trial evidence; weak for RAD21-targeted efficacy | ClinicalTrials.gov entries NCT06789783, NCT05829668, NCT04381897, NCT03113877 (pqac-00000018, pqac-00000019, pqac-00000020, pqac-00000022) |
| Major evidence gaps | No robust subtype-specific prevalence, incidence, survival, penetrance, or natural-history estimates; limited formal QoL data; no validated biomarkers; no RAD21-specific interventional trials; epigenomic/multi-omic data in CDLS4 remain sparse | Evidence-gap summary from available literature | Krab et al. 2020; Kaur et al. 2023; 2024 review/trials landscape (pqac-00000028, pqac-00000046, pqac-00000006, pqac-00000023) |


*Table: This table summarizes the highest-yield current evidence for RAD21-related Cornelia de Lange syndrome, emphasizing cohort size, phenotype frequencies, recent 2023 variant reports, mechanisms, diagnostics, and the absence of subtype-specific therapies. It is useful as a compact reference for knowledge-base population and evidence grading.*