Rationale: Synovial fibrosis, driven by myofibroblast activation and extracellular matrix remodelling, is fundamental in osteoarthritis (OA) pathogenesis but remains poorly understood due to the spatial heterogeneity of synovial inflammation (synovitis). Accurate molecular endotyping of synovial inflammation is essential for effective treatment of OA given its multifactorial nature, yet it requires integrating multiple layers of information with spatial context due to the significant heterogeneity of the tissue. Methods: In this proof-of-concept study, we leveraged MALDI mass spectrometry imaging to achieve spatial metabolomic maps that complement high-content proteomic profiles. Microflow liquid chromatography was employed to improve the robustness and throughput of spatial proteomics. By coupling these spatially resolved datasets, we establish a pseudo time trajectory of heterogeneous synovitis in human knee OA using an integrative framework of spatially resolved proteomics and matrix-assisted laser desorption/ionization mass spectrometry imaging. Results: Clustering 3534 proteins and 79 energy metabolites from spatial proteomic and metabolomic image datasets reveals four distinct functional stages of OA synovitis, i.e., quiescent, microvasculopathic, pre-fibrotic, and post-fibrotic stages, which enables construction of a corresponding pseudo time. Network analyses elucidate the functional links among these stages, highlighting an immune-metabolic axis from endothelial injury and microvascular thrombosis toward myofibroblast activation. Conclusions: This integrative multi-omics imaging approach informs the inflammatory endotype of OA, supporting a vascular aetiology of synovial fibrosis and offering mechanistic insights that could inform more targeted therapeutic strategies. Validation in larger, stratified patient cohorts will be critical to refine our findings and accelerate their clinical usages.