Page 142 - 南京医科大学自然版

第46卷第6期
·924 · 南 京 医 科 大 学 学 报 2026年6月

表2 葡萄膜炎的MR分析总体框架
Table 2 Overall framework for Mendelian randomization analysis of uveitis
Biological mechanism/Clinical
Method type Description/Example Key findings/Applications
significance
Data sources & GWAS databases： IEU Instrumental variables primarily derived Provides reliable genetic variants for
Instruments OpenGWAS，FinnGen，UK from European populations；population MR analysis，ensuring validity and rep⁃
Biobank；Selection criteria：heterogeneity requires attention；future resentativeness of instruments
P < 0.001，LD clumping studies should incorporate East Asian data
+
+
Univariate/Two ⁃ Immune cell traits（e.g.，NK cells and CD14 CD16 monocytes pos⁃ Immune cell subsets involved in pro⁃in⁃
sample MR: im⁃ NK cells，CD14 + CD16 + itively associated with uveitis risk flammatory responses，amplifying IL⁃17/
mune cells monocytes）→Uveitis risk TNF inflammatory pathways；suggests
immunosuppressive targets
Univariate/Two ⁃ Vitamin D levels→Non⁃in⁃ Genetically predicted lower 25⁃OHD lev⁃ Supports individualized vitamin D sup⁃
sample MR: vita⁃ fectious uveitis/scleritis risk els significantly increase uveitis risk；vita⁃ plementation strategies，though dosage
min D min D deficiency enhances Th1/Th17 and population differences require con⁃
pathways sideration
Mediation MR Gut microbiota→Immune Lachnospiraceae abundance negatively Reveals“gut⁃eye axis”mechanism；pro⁃
cells→Acute anterior uve⁃ associated with AAU risk；bacteroides vides rationale for probiotics or fecal
itis（AAU） abundance positively associated；microbi⁃ microbiota transplantation combined
ota influences uveitis via immune cell reg⁃ with immunotherapy
ulation
Multivariable MR Simultaneous inclusion of Ankylosing spondylitis and Crohn’s dis⁃ Controls for inter⁃disease confounding，
（MVMR） ankylosing spondylitis，ease independently associated with in⁃ clarifies shared immune drivers；sug⁃
Crohn’s disease etc. →Iri⁃ creased iridocyclitis risk；Graves’ dis⁃ gests ophthalmologic follow⁃up for high⁃
docyclitis risk ease may be associated with reduced risk risk populations
Bidirectional MR Uveitis ⇄ Systemic immune Validates bidirectional causal relation⁃ Elucidates inter ⁃ disease interactions；
diseases（e.g.，psoriasis，in⁃ ships between uveitis and certain immune provides evidence for comorbidity man⁃
flammatory bowel disease） diseases；e.g.，all subtypes of inflammato⁃ agement
ry bowel disease increase uveitis risk
Drug ⁃ target MR: Probucol→Metabolic path⁃ Probucol reduces Behçet’s disease risk Suggests lipid⁃lowering drugs may exert
Probucol ways→Behçet’s disease⁃re⁃ by inhibiting ABCA1 and lowering HDL therapeutic effects via metabolic ⁃ im⁃
lated uveitis risk levels mune axis；offers clues for drug repur⁃
posing
Drug ⁃ target MR: Glucocorticoid use→Senile Both glucocorticoid use and iridocyclitis Informs clinical decision ⁃ making re⁃
glucocorticoids cataract risk（with iridocy⁃ increase senile cataract risk garding ocular complications when pre⁃
clitis as mediator） scribing glucocorticoids
Multi⁃omics inte⁃ Integration of genomics，Identifies causal relationships between in⁃ Reveals immune ⁃ metabolic interaction
grated MR metabolomics，and micro⁃ flammatory cytokines（e.g.，IL ⁃ 12β，IL ⁃ networks；discovers novel biomarkers
biomics data to construct 33），metabolites（e.g.，mannitol），and and therapeutic targets
causal networks Behçet’s syndrome
Future directions Network MR，colocalization Moving from single causal chains to multi⁃ Enhances robustness and clinical utili⁃
analysis，cross ⁃ population dimensional causal network analysis；fa⁃ ty of causal inference；guides precision
validation，clinical transla⁃ cilitating early screening，drug target pri⁃ medicine strategies
tion oritization，and personalized medicine

MR 分析联合多组学分析。多变量 MR 也是 MR 的
5 未来发展方向
最新扩展，它使用与暴露相关的多个可能的遗传变
MR 分析在发展中可以整合不同的组学，进行 异来估计每次暴露对单个结果的影响 ［36］ 。而基于