Why anti-inflammatory treatments fail in rheumatoid arthritis

In 2018, Rockefeller University rheumatologist Dana Orange was investigating gene expression in the synovium, the soft tissue lining the joints, in patients with rheumatoid arthritis. She identified two subsets of patients: one with synovial inflammation and high pain levels, and another with little inflammation yet significant pain (1). “This was surprising because it had been assumed that pain in rheumatoid arthritis was due to synovial inflammation,” Orange said.

In a 2024 study published in Science Translational Medicine, Orange and her colleagues at Weill Cornell Medicine explored the possible causes of pain in people with rheumatoid arthritis but low synovial inflammation (2). Using machine learning, they uncovered 815 genes associated with nerve growth in 39 such patients. These expression signatures could lead to new therapeutic strategies for the 20 percent of people with rheumatoid arthritis who do not respond to anti-inflammatory treatments but who still experience significant discomfort in their joints (3).

“This is an area of unmet need in terms of therapeutics to address pain when inflammatory pathways are not active,” said Adam Croft, a rheumatologist at the University of Birmingham who was not involved with the study. “This opens up the possibility that there may be pathways that can be modulated to specifically address this problem.”

We cannot say that is the cause of pain [in patients with low synovial inflammation], but it's a smoking gun.
– Dana Orange, Rockefeller University

Once Orange’s team identified genes that may be related to pain in the absence of inflammation, they wanted to know which cell types expressed them. Most of the 815 genes that correlated with pain were particularly active in the lining layer synovial fibroblasts. These cells normally protect the joints from inflammation, but they contribute to joint damage under inflammatory conditions such as rheumatoid arthritis (4).

To understand whether lining layer synovial fibroblasts contribute to pain, the researchers analyzed fibroblast-derived molecules to see if they bound to receptors in the neurons of the dorsal root ganglion, a hotspot for nerve cells that relay pain information to the brain. They found that neuronal receptors bound fibroblast-derived ligands such as netrins, ephrins, integrins, and semaphorins, which signal through pathways known to produce pain. In vitro experiments showed that one such ligand, netrin-4, enhanced the growth of cultured neurons from adult mice.

Additionally, in vivo observations in patients with low-inflammatory synovium showed that fibroblasts expressing pain markers lined areas containing synovial papillary hypertrophy, an abnormal growth of the joint synovial tissue that thickens the synovial membrane's structure and contributes to joint stiffness and reduced mobility. These areas were also connected to blood vessels surrounded by nerve cells that transmit pain. “We cannot say that is the cause of pain [in patients with low synovial inflammation], but it's a smoking gun,” Orange said.

The influence of synovial fibroblasts on nerve development isn't limited to rheumatoid arthritis. “[Orange’s findings] may have implications for other types of inflammatory arthritis,” said Croft. These include osteoarthritis in which the lining layer synovial fibroblasts also enhance neuronal survival and branching (5).

In ongoing work, Orange’s research group is investigating the mechanisms of how lining fibroblasts increase neuronal sprouting. When they cultured neurons in conditioned medium from synovial fibroblasts from rheumatoid arthritis mice, they observed increased neuronal branching, suggesting that multiple factors from synovial fibroblasts influence neuronal development. “We have a list of other possible factors that are predicted to interact with sensory nerves, and we are figuring out which ones might be amenable to [therapeutic] targeting,” she said.