| domain | strongest finding | evidence type | confidence/caveat |
|---|---|---|---|
| Historical gene-disease association | DFNA2B was historically attributed to **GJB3/connexin 31** based on **two small Chinese autosomal-dominant families** with high-frequency hearing loss; later reviews note the evidence base is limited and DFNA2 is genetically heterogeneous. (pqac-00000013, pqac-00000005) | Human family reports summarized in reviews | **Low-moderate confidence** for a historical association; small pedigrees and later contradictory/heterogeneous DFNA2 data limit certainty. |
| Core phenotype | Reported DFNA2B phenotype is **adult-onset**, **progressive**, **sloping/high-frequency** sensorineural hearing loss; a 2023 review states hearing loss may be **milder in females**. (pqac-00000007, pqac-00000012, pqac-00000013) | Human clinical reviews | **Moderate confidence** for the broad audiophenotype; **low confidence** for sex effect because it is sparsely documented in secondary summaries. |
| Functional variant evidence | In vitro studies summarized in a mechanistic review report **V27M, V43M, and V84I** can reach the membrane yet show **loss of dye/ion coupling**, supporting impaired gap-junction function. (pqac-00000009) | In vitro cell-based functional assays (secondary summary) | **Moderate confidence** that these variants can disrupt coupling in cells; direct disease causality in humans remains less certain. |
| Trafficking mechanism | **V174M** was reported to **fail plasma-membrane targeting** and instead accumulate in **lysosomes**; it may also disturb wild-type Cx26 trafficking. (pqac-00000015) | In vitro transfected-cell study (secondary summary) | **Moderate confidence** for a trafficking-defect mechanism; relevance to penetrance and phenotype in vivo is unresolved. |
| Animal-model calibration | **Gjb3/Cx31-null mice do not show hearing impairment**, despite human variant-based disease claims. (pqac-00000015) | Mouse knockout model | **Important caveat / lowers confidence** in a simple haploinsufficiency model; dominant-negative, species-specific, or developmental-context mechanisms remain possible. |
| Cohort/population evidence | In a **2018 Chinese cohort** of 100 unrelated NSHI families, **one** putatively relevant **p.V84I** finding was detected; the paper estimated a very low allele frequency and interpreted the case as **digenic/tri-allelic GJB2/GJB3 ambiguity**, not clean monogenic proof. (pqac-00000008) | Human cohort + segregation/NGS follow-up | **Low-moderate confidence** for contribution of p.V84I; the same study emphasizes ambiguity and possible additive rather than standalone causation. |
| Current therapy landscape | **No DFNA2B/GJB3-specific approved therapy** was identified; management follows general hereditary hearing-loss care (audiology, hearing aids/cochlear implantation as indicated, counseling). Recent precision-diagnostics literature supports broad NGS-based diagnosis rather than gene-specific treatment. (pqac-00000011, pqac-00000007) | Recent clinical review + diagnostic cohort | **High confidence** that no disease-specific therapy currently exists; treatment evidence is extrapolated from broader hereditary hearing-loss practice. |
| Trial/implementation status | The only retrieved GJB3-relevant registered study was **NCT06133946**, an **observational newborn deafness-gene screening cohort**, not an intervention trial; it includes a single GJB3 variant among screened targets. (pqac-00000010) | ClinicalTrials.gov observational study | **High confidence** that current registered activity is screening/epidemiologic rather than therapeutic. |


*Table: This table summarizes the strongest currently retrievable evidence for the historical DFNA2B–GJB3 association, highlighting where support comes from human families, cell studies, mouse models, and modern screening data. It is useful because the evidence is mixed and requires careful confidence calibration rather than a simple yes/no interpretation.*