| Domain | Validated data or candidate annotation | Evidence tier | Caveat |
|---|---|---|---|
| Disease model / genetics | TBI is externally acquired, not Mendelian. **No monogenic etiology, causal pathogenic variant, inheritance pattern, carrier frequency, or clinically indicated TBI genetic test is established.** Lipsky et al., 2024; DOI: [10.3389/fnins.2024.1446076](https://doi.org/10.3389/fnins.2024.1446076). (pqac-00000021, pqac-00000036) | **Established disease-level conclusion** | Do not populate causal-gene or pathogenic-variant fields. *APOE*, *BDNF*, *COMT*, and *GRIN2A* are possible modifiers or exploratory candidates, not diagnostic genes. |
| Epidemiology | GBD 2019 estimated **27.16 million incident cases**, age-standardized incidence of **346 per 100,000**, prevalence of **599 per 100,000**, and 7.08 million YLDs in 2019. Huang et al., published 18 March 2024; DOI: [10.1016/j.cjtee.2024.03.007](https://doi.org/10.1016/j.cjtee.2024.03.007). (pqac-00000011) | **Modeled global population evidence** | GBD estimates are modeled rather than directly enumerated and depend on source coverage, case definitions, and statistical assumptions. |
| Persistent phenotype | In TRACK-TBI, **53%** of participants with TBI reported at least three symptoms at 12 months versus **24%** of orthopedic-trauma controls; the prospective analysis included 2,039 participants with TBI. Machamer et al., 2022; DOI: [10.1089/neu.2021.0348](https://doi.org/10.1089/neu.2021.0348). (pqac-00000012) | **Prospective multicenter human cohort** | Participants were treated at level-1 trauma centers and had CT ordered. Symptom endorsement is not TBI-specific and may not generalize to community mTBI. |
| Single-cell transcriptomics | A murine atlas profiled **334,376 cells** and identified **23 transcriptionally distinct populations** across repetitive closed-head injury, controlled cortical impact, and controlled cortical impact with hemorrhagic shock. Jha et al., 2024; DOI: [10.1016/j.neuron.2024.06.021](https://doi.org/10.1016/j.neuron.2024.06.021). (pqac-00000014, pqac-00000039) | **High-quality preclinical discovery evidence** | Mouse models do not reproduce the full heterogeneity of human TBI. Cell states and candidate targets require spatial, functional, and human validation. |
| Blood biomarkers | Among **130 adults** with nonpenetrating mTBI and GCS 13–15 tested within 12 hours, CT identified injury in only **7 patients (5%)**. Combined GFAP/UCH-L1 had sensitivity 1.00 and NPV 1.00; 96 tests were positive and 34 negative. Legramante et al., 2024; DOI: [10.1186/s12245-024-00708-z](https://doi.org/10.1186/s12245-024-00708-z). (pqac-00000016) | **Single-center retrospective diagnostic study** | Seven CT-positive cases yielded wide sensitivity uncertainty (95% CI 0.64–1.00). The test supports CT triage but does not independently diagnose or exclude every form of TBI. |
| Surgery | RESCUEicp randomized 408 patients with refractory ICP above 25 mm Hg. Six-month mortality was **26.9% with decompressive craniectomy versus 48.9% with medical care**, but surgery increased survival with vegetative state or severe disability; adverse events were 16.3% versus 9.2%. Hutchinson et al., 2016; DOI: [10.1056/NEJMoa1605215](https://doi.org/10.1056/NEJMoa1605215). (pqac-00000033) | **Multicenter randomized controlled trial** | This was rescue treatment for selected refractory intracranial hypertension, not routine care. Survival benefit must be weighed against disability, complications, and patient goals. |
| Pharmacotherapy / disorders of consciousness | In **184 patients** in a vegetative or minimally conscious state 4–16 weeks after TBI, amantadine accelerated recovery during four weeks of treatment: Disability Rating Scale slope difference **0.24 points per week** versus placebo (*P*=0.007). Giacino et al., 2012; DOI: [10.1056/NEJMoa1102609](https://doi.org/10.1056/NEJMoa1102609). (pqac-00000026) | **Multicenter double-blind randomized controlled trial** | Benefit concerned recovery rate during active treatment; overall improvement was similar by week 6 after washout. Results apply to prolonged traumatic disorders of consciousness, not uncomplicated mTBI. |
| Genetic susceptibility | A veteran GWAS included 111,494 TBI cases and 192,991 controls and reported 15 genome-wide significant loci and 14 gene-level signals, including *NCAM1*, *APOE*, *FTO*, and *FOXP2*. Merritt et al., summarized by Lipsky et al., 2024; DOI: [10.3389/fnins.2024.1446076](https://doi.org/10.3389/fnins.2024.1446076). (pqac-00000021) | **Large human association study; candidate annotation** | Signals may reflect exposure propensity or risk-taking rather than biological susceptibility to tissue injury. Significant SNP findings were confined to European-ancestry participants and require diverse replication. |


*Table: Concise knowledge-base annotations spanning epidemiology, phenotypes, biomarkers, omics, genetics, surgery, and pharmacotherapy. Evidence tiers and caveats distinguish validated clinical findings from modeled estimates and candidate associations.*