| Topic | Key finding/statistic | Population/study design | Publication (authors, journal) | Year/month | DOI/URL | Evidence type |
|---|---|---|---|---|---|---|
| Diagnostic criteria (ICON/International Consensus) | Clinical criteria: A = monocular, subacute vision loss with orbital pain on eye movement, reduced contrast/color vision, RAPD; B = painless with all other features of A; C = binocular loss with features of A or B. Paraclinical thresholds: OCT mGCIPL inter-eye difference **>4% or >4 μm** or pRNFL **>5% or >5 μm** within **≥3 months** after onset; MRI symptomatic optic nerve/sheath contrast enhancement acutely or intrinsic T2 signal increase within **≥3 months**; biomarkers: **AQP4, MOG, CRMP5 seropositivity or CSF oligoclonal bands**. Definite ON: A + 1 paraclinical test; B + 2 tests of different modality; C + MRI + 1 other paraclinical test (pqac-00000023) | International Delphi consensus diagnostic/classification framework for optic neuritis | Petzold et al., *Lancet Neurology* | 2022/Sep | 10.1016/S1474-4422(22)00200-9; https://doi.org/10.1016/S1474-4422(22)00200-9 | Guideline/consensus (human expert Delphi) |
| OCT thresholds validated in AQP4+ NMOSD-ON | Proposed thresholds tested: pRNFL IEAD **5 μm**, IEPD **5%**; GCIPL IEAD **4 μm**, IEPD **4%**. NMOSD-ON vs healthy controls: pRNFL IEAD **AUC 0.95, specificity 82%, sensitivity 86%**; GCIPL IEAD **AUC 0.93, specificity 98%, sensitivity 75%**; pRNFL IEPD **AUC 0.96, specificity 87%, sensitivity 89%**; GCIPL IEPD **AUC 0.94, specificity 96%, sensitivity 82%**. NMOSD-ON vs NMOSD-NON: pRNFL IEAD **AUC 0.92, specificity 77%, sensitivity 86%**; GCIPL IEAD **AUC 0.87, specificity 85%, sensitivity 75%**; pRNFL IEPD **AUC 0.94, specificity 82%, sensitivity 89%**; GCIPL IEPD **AUC 0.88, specificity 82%, sensitivity 82%** (pqac-00000016) | Multicenter study of **28** AQP4+ NMOSD unilateral ON, **45** AQP4+ NMOSD without ON history, **62** healthy controls | Oertel et al., *JNNP* | 2023/Feb | 10.1136/jnnp-2022-330608; https://doi.org/10.1136/jnnp-2022-330608 | Human multicenter diagnostic cohort |
| OCT thresholds validated in MOG-ON | Using ICON-related published cutoffs (>4% or >4 μm mGCIP; >5% or >5 μm pRNFL), pooled MOG-ON analysis showed mGCIP IEPD sensitivity **92%**, mGCIP IEAD **88%**, pRNFL **84%**; specificity: mGCIP IEPD **82%**, mGCIP IEAD **82%**, pRNFL IEPD **82%**, pRNFL IEAD **79%**. In unilateral ON, diagnostic sensitivity was **>99% for all metrics**; in bilateral ON, sensitivity fell to **61%–78%** (pqac-00000019) | Multicenter validation study of **66** participants (**33 MOG-ON**, **33 controls**) | Volpe et al., *Neurology Neuroimmunology & Neuroinflammation* | 2024/Nov | 10.1212/NXI.0000000000200291; https://doi.org/10.1212/NXI.0000000000200291 | Human multicenter diagnostic cohort |
| PLEX outcomes in severe ON | **395** ON attacks in **317** patients; median time from vision loss onset to PLEX **2.6 weeks** (IQR **1.4–4.0**); median VA at PLEX **count fingers** (IQR **20/200–hand motion**); median final VA **20/25** (IQR **20/20–20/60**); **81 attacks (20.5%)** ended with final VA **20/200 or worse**. Poorer outcomes associated with older age (**p=0.002**), worse VA at PLEX (**p<0.001**), and longer delay to PLEX (**p<0.001**). In matched comparison, final VA worse than 20/40 occurred in **12% (6/50)** of PLEX-treated attacks vs **33% (6/18)** in ONTT IVMP without PLEX (**p=0.04**) (pqac-00000007) | International multicenter retrospective analysis; etiologies included MS (**108**), MOGAD (**92**), AQP4+ NMOSD (**75**), seronegative NMOSD (**34**), idiopathic (**83**) | Chen et al., *American Journal of Ophthalmology* | 2023/Aug | 10.1016/j.ajo.2023.02.013; https://doi.org/10.1016/j.ajo.2023.02.013 | Human multicenter retrospective cohort |
| MOGAD / MOG-ON epidemiology and phenotype | MOGAD incidence **~1.6–4.8 per million/year**; prevalence **1.3–2.5 per 100,000**; biphasic onset peaks at **5–10 years** and **20–45 years**. Optic neuritis is the commonest adult presentation (**~30–60%**). Distinctive features include frequent bilateral onset, optic disc swelling, headache, longitudinally extensive optic nerve hyperintensity, and optic perineuritis on MRI; serum live cell-based assay MOG-IgG is highly specific (pqac-00000018) | Narrative review synthesizing cohort/registry data on MOG antibody-associated optic neuritis | Jeyakumar et al., *Eye* | 2024/May | 10.1038/s41433-024-03108-y; https://doi.org/10.1038/s41433-024-03108-y | Human review of cohort/registry evidence |
| NMOSD prevalence in the US | 2022 prevalence **6.88/100,000** from **1,772** patients among **25,743,039** EHR patients. By race: Black **12.99/100,000**, Asian **9.41/100,000**, White **5.58/100,000**. Among females, prevalence **9.48/100,000**; Black and Asian females had **2.65×** and **1.94×** higher prevalence than White females. Male prevalence **3.52/100,000**; female:male ratio about **3.5:1**. Estimated US burden: **15,413 females** and **6,233 males** (~**22,000** total) (pqac-00000024) | Cross-sectional prevalence study using nationwide aggregated EHR data from 55 healthcare organizations | Briggs & Shaia, *Multiple Sclerosis Journal* | 2024/Jan | 10.1177/13524585231224683; https://doi.org/10.1177/13524585231224683 | Human population-based EHR epidemiology |
| MS risk stratification after ON using genetics | In undifferentiated ON, one SD increase in MS genetic risk score (MS-GRS) increased hazard of MS **1.3-fold**; multivariable HR **1.29** (95% CI **1.07–1.55**, **P=0.0067**). Predicted-risk quartiles developed incident MS at rates from **4%** (95% CI **0.5–7%**, lowest quartile) to **41%** (95% CI **33–49%**, highest quartile). Background rates in study: ON incidence cited as **3.7–5.1 per 100,000 person-years** in UK/USA; by 5 years, about **20%** of undifferentiated ON diagnosed with MS; by 15 years, up to **50%** of ON (excluding bilateral presentation) diagnosed with MS. In UK Biobank, among **545** undifferentiated ON cases, **124 (22.8%)** developed MS during median **18.4 years** follow-up (pqac-00000025, pqac-00000026) | UK Biobank primary analysis with replication in Geisinger and FinnGen; genetic risk modeling in ON | Loginovic et al., *Nature Communications* | 2024/Feb | 10.1038/s41467-024-44917-9; https://doi.org/10.1038/s41467-024-44917-9 | Human genomic cohort / prognostic modeling |


*Table: This table compiles high-yield quantitative findings from 2022-2024 on optic neuritis diagnostic criteria, imaging thresholds, outcomes, epidemiology, and MS risk prediction. It is useful as a compact evidence summary for knowledge-base curation and citation-backed reporting.*