Primary open-angle glaucoma (POAG) is a complex eye disease characterized by progressive loss of optic nerve function that, if untreated, ultimately leads to irreversible blindness. To date, the biological mechanisms causing POAG are still unclear. There is disparity in POAG prevalence, clinical presentations and outcomes across ancestries. Here, we aim to identify unique genetics that underlie risk to POAG and evaluate the potential connection with vascular mechanisms. We performed POAG meta-analysis across 15 biobanks that are part of the Global Biobank Meta-analysis Initiative, with two previously published multi-ancestry analyses for a total of 1,478,037 individuals from six ancestries (46,325 cases and 1,431,712 controls). A total of 109 genome-wide significantly associated loci (p<5e-8) were identified, 18 of which were novel. Three of these novel loci are ancestry-specific, two African-specific and the third specific to northern Europeans. We also identified five sex-specific novel loci, four of which are African-specific and one European-specific. To explore biological implications underlying these variant-trait associations, we performed gene enrichment analysis, gene prioritization analysis and transcriptome-wide association studies (TWAS) implicating genes related to vascular-related functions, blood vessels, angiogenesis, and cancer. A fifth of TWAS-prioritized genes with vascular-related and/or cell senescence/proliferation functional roles or have been implicated in vascular or neoplastic diseases are primary ciliary related genes. We further performed extensive statistical validation analysis of genes in the SIX6 and well-known CDKN2B-AS1 loci, previously implicated in POAG, cardiovascular diseases, and cancers across multiple ancestries. We found evidence of significant interaction between SIX6 rs33912345 and causal variants in chr9p21.3, with concomitant effect on expression of a primary cilia gene CDKN2A and CDKN2B at the CDKN2B-AS1 locus. Phenome-wide association analysis of POAG genetic risk burden across five biobanks and genetic correlation analysis also show the shared biology between POAG and vascular and neoplastic traits. In summary, our findings suggest that some POAG risk variants may be ancestry-specific, sex-specific, or both. Our findings further support the contribution of vascular and proliferation genes in POAG and suggest potential involvement of primary cilia in POAG pathogenesis.