Does vascular endothelial growth factor play a role in interleukin-6 receptor antagonist therapy for rheumatoid arthritis?


The age of anticytokine therapy

It is hard to recall the days when anticytokine therapy was just an interesting hypothetical concept. Due to the spectacular success of anti–tumor necrosis factor (anti-TNF) therapy, it now appears the sky is the limit for identification of potential new targets whose antagonism may prove beneficial in the treatment of rheumatoid arthritis (RA). The effectiveness of anti-TNF in the treatment of RA has clearly demonstrated that systemic modulation of “cytokines” can produce profound beneficial effects on a localized pathogenic process (1). The medical community owes a huge debt of gratitude to Marc Feldmann, Ravinder Maini, and many other investigators, for their vision and work over several decades that resulted in the new age of anticytokine therapy in the clinic. The initial identification of interleukin-6 (IL-6) as a promising candidate for anticytokine therapy in the treatment of RA (2) is confirmed by Nakahara and colleagues in a report published in this issue of Arthritis & Rheumatism (3). Moreover, those investigators have begun to explore the mechanisms that may underlie the efficacy of IL-6 antagonism in RA and demonstrate that IL-6/IL-6 receptor (IL-6R) complexes, together with IL-1, can induce vascular endothelial growth factor (VEGF) production by synovial fibroblasts.

IL-6 antagonism as a target in RA

IL-6 has been studied in a variety of settings over many years and is a good example of a multipotent cytokine. IL-6 has been shown to be involved in several biologic processes that could be important to sustaining chronic synovitis. These processes include inflammation, plasma cell differentiation, megakaryopoiesis/platelet production, and others (4). Following the pathway used for identification of anti-TNF therapy (5), experiments with the collagen-induced arthritis model were first used to demonstrate the efficacy of IL-6 neutralizing antibodies (6). Later, a humanized monoclonal antibody directed against IL-6R was shown to be effective at reducing the signs and symptoms of arthritis in the first anti–IL-6R clinical trial (2). Nakahara et al now show, in a followup study (3), that systemic IL-6R antagonism with a humanized monoclonal antibody decreases the signs and symptoms of arthritis as evaluated by American College of Rheumatology 20% improvement (ACR20) and ACR50 criteria (7, 8). The authors have also begun to explore possible molecular mechanisms for the efficacy of anti–IL-6 therapy. Specifically, they show that antagonism of IL-6 activity in patients reduces the elevated serum levels of VEGF typical of arthritis and also demonstrate that IL-6, together with IL-1, augments VEGF production in primary cultures of human synovial fibroblasts.

Synergism in regulation of VEGF secretion by synovial fibroblasts

Local production of VEGF in arthritic synovial tissue has been documented (9) and appears to correlate with disease activity in humans. The factors that regulate VEGF production by synovial fibroblasts and other cells in arthritic synovial tissue are therefore of great interest, especially since VEGF has been shown to be important in arthritis pathogenesis in animal models of RA (10–12). Nakahara and colleagues show in their current report (3) that there is marked synergism between IL-6 and IL-1 in the induction of VEGF by IL-6/IL-6R complexes in cultured synovial cells. It should be noted that IL-6 differs from most cytokines in the manner by which it signals to target cells. IL-6 utilizes a soluble receptor (sIL-6R) which binds both IL-6 and another protein called gp130, to transmit signals to responding cells (13). Therefore, in the experimental culture system used by Nakahara et al, exogenous IL-6R was used. It is very likely, however, that there is sufficient sIL-6R present in inflamed synovium to allow IL-6/sIL-6R complexes signaling via gp130, together with IL-1 and perhaps other cytokines, to contribute to the production of VEGF in intact arthritic synovial tissue.

Synergistic regulation of VEGF has also been reported with other cytokines (i.e., transforming growth factor β and IL-1) (14). This suggests that to fully understand how synovial cells behave in vivo, we must acknowledge their exposure to a complex cytokine milieu as well as hypoxia in arthritic synovial tissue (14–16).

Reduction of serum VEGF and C-reactive protein levels by IL-6R antibody therapy

Nakahara and colleagues show that elevated serum levels of 2 indicator proteins associated with chronic synovitis, i.e., C-reactive protein and VEGF, were reduced to normal in RA patients after 8 weeks and 24 weeks of anti–IL-6 therapy. Elevated levels of VEGF in the serum of RA patients have been reported previously (17) and examined in depth by Paleolog and her colleagues at the Kennedy Institute of Rheumatology in London. Similar to the findings reported now by Nakahara et al, Paleolog et al showed that anti-TNF therapy caused a rapid reduction of VEGF levels in RA patient serum (17). A more recent study reported in Arthritis & Rheumatism demonstrated that elevated serum levels of VEGF occur in multiple forms of arthritis. The highest amounts of VEGF were found in the early phase of RA, and serum VEGF levels correlated well with disease progression (18).

While both anti-TNFα and anti–IL-6R therapies most likely modulate more than one process important in the pathogenesis of RA, here we will consider one potential ramification of the modulation of serum levels of VEGF. We will discuss the possibility that VEGF may serve as a prototype for cytokines that could be transported to the bone marrow compartment via the circulation. Cytokines in the circulation might subsequently influence the homeostasis of hematopoietic progenitor pathways that support inflammation (inflammatory effector cells) (12) and angiogenesis (endothelial progenitor cells) (19, 20).

“Systemic VEGF” and its importance in arthritis pathogenesis

Because inflamed synovium is most often both a source of cytokines and the potential target of the actions of many cytokines, we have primarily concentrated on synovium and other joint tissues in defining mechanisms by which cytokines affect arthritis. New findings suggest we may have overlooked additional consequences of “inflammatory” or “angiogenic” cytokines, which may be generated in synovium but can find their way into the circulation and exert important biologic effects remote from joint tissues (Figure 1). Specifically, recent reports provide evidence that VEGF and placenta growth factor (PlGF), which also can bind Flt-1, may affect arthritis pathogenesis through their actions on multipotent and committed hematopoietic progenitor cells in the bone marrow which express one of the VEGF receptors, either Flk-1 or Flt-1 (12, 21).

Figure 1.

“Systemic” effects of cytokines on disease. VEGF = vascular endothelial growth factor; GM-CSF = granulocyte–macrophage colony-stimulating factor; IL-6 = interleukin-6.

Effects on arthritis via VEGF receptors on bone marrow progenitors

The concept that VEGF targets only differentiated endothelial cells and monocytes has changed recently. Studies have revealed powerful effects of VEGF and its family member PlGF (22) on hematopoietic pathways (23–25) that generate endothelial progenitor cells and inflammatory cells (12), both of which may have important roles in disease pathogenesis in the synovium. A novel progenitor cell for endothelial cells, whose lineage is regulated via VEGF receptor Flk-1, appears to be a very important component of “angiogenesis” in adults, and evidence indicates that elevation of circulating VEGF increases the mobilization of these cells from the bone marrow (19, 25, 26). Endothelial progenitor cells incorporate into newly forming vessels (20), although their exact function in vascular morphogenesis is presently unknown (20). A recent report in the New England Journal of Medicine has linked reduced numbers of circulating endothelial progenitors with increased risk of heart disease, suggesting a role in maintenance of vascular function (27). This report may be just the beginning of the recognition of the importance of this cell type in disease states and the cytokines that regulate its production.

The development of 2 inhibiting monoclonal antibodies (12, 28) made it possible to evaluate the specific contribution of each of the 2 predominant VEGF receptors (VEGFR-1, or Flt-1 and VEGFR-2, or kinase insert domain receptor/Flk-1) in animal models of arthritis and other disorders in which angiogenesis and inflammation play important roles (12). A multicenter study using these unique VEGF receptor–specific monoclonal antibodies resulted in startling experimental results that may forever alter our way of viewing the involvement of cytokines in arthritis and other diseases (12). These studies showed that reduced severity of arthritis in a murine model could be achieved by antagonism of only the VEGF receptor Flt-1, through a previously unknown pathway that regulates the generation and mobilization of inflammatory effector cells in the bone marrow compartment (12).

The seminal studies performed with these reagents demonstrated that inhibition of only 1 of the VEGF receptors, Flt-1, but not Flk-1, attenuated arthritis in a murine model of RA. The key to understanding the unexpected efficacy of Flt-1 antagonism was the discovery of its role in controlling the numbers of inflammatory cells (monocytes, granulocytes) that arise in bone marrow and support the inflammatory reaction in the joints (12). Thus, VEGF and PlGF, the ligands for Flt-1, can affect inflammation by regulating the abundance of inflammatory effector cells. This was an unexpected finding, since most previous studies on VEGF indicated that angiogenesis was driven by VEGF through the receptor Flk-1. Previous in vivo studies used reagents that antagonized VEGF activation of both Flk-1 and Flt-1, and it was assumed that beneficial effects of VEGF antagonism resulted from reduced Flk-1–driven angiogenesis in animal models. Therefore, it was somewhat unsettling when in vivo inhibition of Flk-1 with a monoclonal antibody failed to ameliorate arthritis (12). In this regard, the lack of a beneficial effect of Flk-1 antagonism on murine arthritis does not preclude the possibility that the same approach may be successful in human arthritis. The more protracted synovial hyperplasia of chronic arthritis in humans may have a larger “angiogenic” component in its pathogenesis.


Nakahara and colleagues (3) have shown IL-6R antagonism to be another anticytokine therapy with promise for the treatment of RA. The molecular mechanisms that underlie the efficacy of IL-6R antagonism are likely complex and include actions on several important cell types that participate in the pathogenesis of arthritis. One possibility may be effects on VEGF production and its potential downstream effects on cells that support synovitis and the associated tissue degradation. It will be very exciting to watch as future studies investigate the role of VEGF induction by IL-6 in the reduction of clinical severity of arthritis and unravel the exact mechanisms by which IL-6R antagonism elicits its beneficial action.