Clinical Investigation of the Role of the Lymphatic Vessels in Regulating Synovial Inflammation in Osteoarthritis

Abstract

Osteoarthritis (OA) is the most prevalent chronic joint disease and is characterized by cartilage erosion, remodeling of the subchondral bone, and persistent inflammation of the synovial membrane (synovitis). Beyond its structural and degenerative aspects, OA is increasingly recognized as an inflammatory disorder in which the vasculature plays a key role in regulating the persistence or resolution of synovitis. While the contribution of blood vessels to OA pathophysiology has already been relatively well studied, the role of lymphatic vessels remains comparatively underexplored, despite their potential importance in modulating fluid clearance and immune cell trafficking within the synovium. Defective lymphatic drainage has been hypothesized to promote local accumulation of inflammatory mediators, thereby sustaining disease progression. Nonetheless, the precise contribution of synovial lymphatic vessels to OA pathophysiology remains poorly understood.

The aim of this master's thesis was to characterize alterations in the vascular networks of osteoarthritic synovial membranes, with a specific focus on lymphatic vessels. To this end, a multimodal strategy was implemented, combining histological staining, immunohistochemistry (IHC), immunofluorescence, quantitative image analysis, transcriptomic profiling, and functional indirect co-culture assays.

Patient-derived human synovial sections from inflamed (I) and normal reactive (NR) regions were stained with both Masson’s trichrome and immunofluorescence (DAPI, PDPN, PROX1, CD31), enabling discrimination between blood and lymphatic vasculature inside the synovial subintima. A dedicated semi-automated and objective image analysis pipeline for vessel quantifications was developed to segment the subintima, align multimodal images, and quantify vessel density and mean vessel size in the subintima. Pipeline robustness was confirmed by quantitative error estimates at specific stages of the code. The automated quantitative analysis revealed that blood vessel density is consistently higher or equal in I regions compared to NR regions. In contrast, lymphatic vessels were detected in fewer samples than visually anticipated. Careful inspection indicated that this under-detection was mainly due to technical limitations, such as suboptimal staining of PDPN or PROX1 and discontinuity of fragile lymphatic structures, which were not accounted for by the quantification pipeline. Nevertheless, some patient samples displayed lymphatic vessels in both regions, with reduced density and smaller mean size in inflamed tissues, suggesting a potentially defective lymphatic response to inflammation.

To complement the spatial information provided by histology, transcriptomic profiling was performed on patient-derived synoviocytes isolated from the I and NR regions. Expression levels of pro-inflammatory (IL-6) and pro-vascular (VEGF-A, VEGF-C) markers were quantified by RT-qPCR. Results demonstrated a heterogeneous but consistent upregulation of IL-6 in inflamed regions, though not systematically matching the surgeon’s initial classification. VEGF-A and VEGF-C expression appeared correlated with IL-6, linking inflammatory activation with vascular remodeling. Additional experiments on commercially available synoviocytes, pre-conditioned or not with inflammatory stimuli, confirmed these trends and validated the experimental pipeline with reproducible in vitro models.

Finally, functional assays using conditioned media concentrated from either synoviocytes or LECs demonstrated reciprocal interactions between these two cell types without direct contact. Inflammatory synoviocyte-derived factors strongly promoted LEC proliferation. Conversely, factors secreted by inflamed LECs enhanced synoviocyte growth. These findings indicate that paracrine signaling may drive stromal expansion and sustain inflammation within the synovial microenvironment, either independently of, or in addition to, direct cell–cell interactions.

Altogether, this work provides new insights into the role of synovial vasculature in OA. It demonstrates the robustness of the developed image analysis pipeline for vessel quantification, while also identifying key limitations and suggesting potential improvements in lymphatic vessel detection. Furthermore, integration of transcriptomic and functional data revealed other outcomes, including a correlation between IL-6 expression and VEGF-A and VEGF-C levels, as well as reciprocal paracrine communication between synoviocytes and LECs. By bridging histological, molecular, and functional perspectives, this study strengthens the hypothesis that impaired lymphatic remodeling could contribute to the persistence of synovial inflammation and represents a promising target for therapeutic strategies in OA.