Anti-inflammatory pharmacotherapy with ketoprofen ameliorates experimental lymphatic vascular insufficiency in mice

Abstract

Background: Disruption of the lymphatic vasculature causes edema, inflammation, and end-tissue destruction. To assess the therapeutic efficacy of systemic anti-inflammatory therapy in this disease, we examined the impact of a nonsteroidal anti-inflammatory drug (NSAID), ketoprofen, and of a soluble TNF-alpha receptor (sTNF-R1) upon tumor necrosis factor (TNF)-alpha activity in a mouse model of acquired lymphedema.

Methods and findings: Lymphedema was induced by microsurgical ablation of major lymphatic conduits in the murine tail. Untreated control mice with lymphedema developed significant edema and extensive histopathological inflammation compared to sham surgical controls. Short-term ketoprofen treatment reduced tail edema and normalized the histopathology while paradoxically increasing TNF-alpha gene expression and cytokine levels. Conversely, sTNF-R1 treatment increased tail volume, exacerbated the histopathology, and decreased TNF-alpha gene expression. Expression of vascular endothelial growth factor-C (VEGF-C), which stimulates lymphangiogenesis, closely correlated with TNF-alpha expression.

Conclusions: Ketoprofen therapy reduces experimental post-surgical lymphedema, yet direct TNF-alpha inhibition does not. Reducing inflammation while preserving TNF-alpha activity appears to optimize the repair response. It is possible that the observed favorable responses, at least in part, are mediated through enhanced VEGF-C signaling.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Tail volume response to pharmacotherapy.

Changes in tail volume are expressed as a percentage of the volume on day 0. By day 7, lymphedema (LY) mice demonstrate significant increase in tail volume. NSAID-treated lymphedema (LY-NSAID) mice were significantly less edematous than normal controls (NL) or sham surgery control (SH) mice. Conversely, sTNF-R1-treated lymphedema (LY-sTNF-R1) mice were significantly more edematous. *P<0.05 vs LY, †P<0.05 vs LY-NSAID.

Figure 2. Histological responses to pharmacotherapy.

Tail sections were harvested 16 mm from the base of the tail, stained with hematoxylin/eosin, and examined by light microscopy. (A) Representative histology. Specimens from normal control (NL) mice, sham surgery control (SH) mice, and lymphedema (LY) mice treated with either PBS, ketoprofen (NSAID), or the TNF-α inhibitor sTNFR-1. Untreated LY show hyperkeratosis, epidermal spongiosis and edema, irregularity of the epidermal/dermal junction, elongation of the dermal papillae, and a 2- to 3-fold expansion of tissue between the bone and the epidermis. There are numerous dilated microlymphatics and increased cellularity, including a large infiltration of neutrophils. Treatment with NSAID normalizes these pathological findings whereas treatment with sTNFR-1 exacerbates the pathology. (B) Quantification of epidermal thickness (ET). Changes in ET are expressed as a percentage of the average ET of NL. ET of NSAID-treated lymphedema (LY-NSAID) mice was significantly reduced compared to untreated LY mice (P<0.0005) and were not significantly different than NL or SH control mice. ET of sTNF-R1 treated lymphedema (LY-sTNF-R1) mice was significantly increased compared to untreated LY mice (P<0.05). *P<0.05 vs LY, †P<0.0005 vs LY-NSAID.

Figure 3. Targeted gene expression analysis by quantitative real-time PCR.

Fold-changes of gene expression are relative to normal (NL) controls on day 11. NSAID treatment significantly induced the expression of TNF-α, MCP-1, VEGF-C, VEGFR-3 and Prox1 in lymphedema (LY) mice. The TNF-α inhibitor sTNF-R1 downregulated TNF-α but did not affect other genes. *P<0.05 vs LY, †P<0.05 vs LY-NSAID.

Figure 4. TNF-α levels in tissue homogenates of tail skin.

Median fluorescence intensity was used to assess the relative tissue concentrations of TNF-α. TNF-α levels were significantly higher in mice with NSAID-treated lymphedema (LY-NSAID) (P<0.005) than in normal controls or mice with untreated or sTNF-R1-treated lymphedema (LY-sTNF-R1). *P<0.005 vs untreated lymphedema, †P<0.005 vs LY-NSAID.

Figure 5. Inflammatory cytokine levels in tissue homogenates of tail skin.

Median fluorescence intensity was used to assess the relative tissue concentrations of MCP-1, MCP-3, MIP1a, Eotaxin, and VEGF-A. MCP-1 levels were elevated in mice with lymphedema (LY) and further increased by both ketoprofen and sTNF-R1 anti-inflammatory treatments. MCP-3 levels were also elevated in lymphedema but were significantly reduced by both treatments. A similar pattern was observed for macrophage inflammatory protein 1a (MIP1a). Like TNF-α, eotaxin levels were reduced by NSAID therapy and decreased by sTNF-R1. No effect on the blood vascular growth factor VEGF-A was observed. *P<0.04 vs to LY, †P = 0.05 vs LY, ‡P<0.05 vs LY-NSAID.

Figure 6. Model of inflammation and lymphedema.

Loss of lymphatic vascular integrity leads to diminished lymph transport, which promotes both edema and inflammation. TNF-α, a potent mediator of inflammation, is also a known inducer of the pro-lymphangiogenic factor, VEGF-C. Both ketoprofen and pegsunercept have general inhibitory effects on inflammation, ketoprofen promotes endogenous repair mechanisms mediated by VEGF-C and VEGFR-3by simultaneous inducing TNF-α. In contrast, pegsunercept directly inhibits TNF-α and therefore exacerbates the disease state by disrupting pro-lymphangiogenic processes driven by VEGF-C and VEGFR-3.