Angiogenic gene expression and vascular density are reflected in ultrasonographic features of synovitis in early Rheumatoid Arthritis: an observational study

Abstract

Introduction: Neovascularization contributes to the development of sustained synovial inflammation in the early stages of Rheumatoid Arthritis. Ultrasound (US) provides an indirect method of assessing synovial blood flow and has been shown to correlate with clinical disease activity in patients with Rheumatoid Arthritis. This study examines the relationship of US determined synovitis with synovial vascularity, angiogenic/lymphangiogenic factors and cellular mediators of inflammation in a cohort of patients with early Rheumatoid Arthritis (RA) patients prior to therapeutic intervention with disease modifying therapy or corticosteroids.

Methods: An ultrasound guided synovial biopsy of the supra-patella pouch was performed in 12 patients with early RA prior to treatment. Clinical, US and biochemical assessments were undertaken prior to the procedure. Ultrasound images and histological samples were obtained from the supra-patella pouch. Histological samples were stained for Factor VIII and a-SMA (a-smooth muscle actin). Using digital imaging analysis a vascular area score was recorded. QT-PCR (quantitative-PCR) of samples provided quantification of angiogenic and lymphangiogenic gene expression and immunohistochemistry stained tissue was scored for macrophage, T cell and B cell infiltration using an existing semi-quantitative score.

Results: Power Doppler showed a good correlation with histological vascular area (Spearman r--0.73) and angiogenic factors such as vascular endothelial growth factor-A (VEGF-A), Angiopoietin 2 and Tie-2. In addition, lymphangiogenic factors such as VEGF-C and VEGF-R3 correlated well with US assessment of synovitis. A significant correlation was also found between power Doppler and synovial thickness, pro-inflammatory cytokines and sub-lining macrophage infiltrate. Within the supra-patella pouch there was no significant difference in US findings, gene expression or inflammatory cell infiltrate between any regions of synovium biopsied.

Conclusion: Ultrasound assessment of synovial tissue faithfully reflects synovial vascularity. Both grey scale and power Doppler synovitis in early RA patients correlate with a pro-angiogenic and lymphangiogenic gene expression profile. In early RA both grey scale and power Doppler synovitis are associated with a pro-inflammatory cellular and cytokine profile providing considerable validity in its use as an objective assessment of synovial inflammation in clinical practice.

Figure 1

Synovial tissue stained with factor VIII demonstrating vessels and gated digital analysis for different sized vessels. (A) Digital image of immunofluorescent-stained synovial blood vessels at 20 times magnification (Olympus BX60 microscope). Synovial tissue with factor VIII staining indicating vessel wall fluorescence (pink). (B) Division of vessel size using digital imaging analysis based on specified generated gates with colour differentiation: blue/green, large; red, medium; grey, small. (C) Corresponding vascular histology stained with vWF - von willebrand factor, CD68 immunohistochemistry and knee power Doppler images for the midline supra-patella pouch (SPP). vWF and CD68 staining at 10 times magnification (Olympus BX60 microscope). Ultrasound (US) images from GE logiq 9 machine with 12 MHz probe. High, intermediate and low levels of vascularity and inflammation indicated.

Figure 2

There was no significant variation of ultrasound, vascular area and gene expression between each region of the supra-patella pouch (SPP). (A) TNF-α expression by each region of interest within the SPP (Kruskal-Wallis test, P = 0.31). (B) Quantitative power Doppler area (PQuant) (pixels) by each region of interest within the SPP (Kruskal-Wallis test, P = 0.27). (C) Synovial thickness quantitative area (SQuant) (pixels) by each region of interest within the SPP (Kruskal-Wallis test, P = 0.19). (D) Synovial vascular area/mm2 by each region of interest within the SPP (Kruskal-Wallis test, P = 0.09). (E) PQuant/SQuant ratio by each region of interest within the SPP (Kruskal-Wallis test, P = 0.44).

Figure 3

Quantitative power Doppler area (PQuant) and synovial thickness quantitative area (SQuant) correlate with angiogenic and lymphangiogenic factors. Graphical representation of relationship of PQuant and SQuant (measured in pixels) with vascular endothelial growth factor (VEGF)α, Angiopoietin 2, IL-1β and VEGF-R3. Regression line and 95% CI shown in each graph. (A) Correlation between PQuant and VEGFa on the y-axis (Spearman r = 0.52). (B) PQuant correlated with Angiopoietin 2 (Spearman r = 0.61). (C) PQuant correlated with IL-1β (Spearman r = 0.68). (D) PQuant correlated with VEGF-R3 (Spearman r = 0.61). (E) SQuant correlated with VEGFα on the y-axis (Spearman r = 0.56). (F) SQuant correlated with Angiopoietin 2, (Spearman r = 0.74). (G) SQuant correlated with IL-1β (Spearman r = 0.81). (H) SQuant correlated with VEGF-R3 (Spearman r = 0.61).

Figure 4

Synovial vascular area predicts Doppler signal within the knee joint. (A) Qualitative Doppler assessment (PQuant) correlates with synovial vascular density within the supra-patella pouch (Spearman r 0.73). (B) PQuant correlates with blood vessel density (Spearman r 0.42). (C) Thresholded Doppler signal for high intensity signal (PDHi) correlates with medium-sized vessel vascular area (Spearman r 0.56).