Summary of studies directly correlating MRI features with synovial pathobiology: histobiological characteristics.
| Study | No. Biopsies | Macroscopic Assessment of Synovium | Routine H&E Assessment | IHC Analysis | Synovial Gene Expression Analysis | Main Conclusion |
|---|---|---|---|---|---|---|
| Konig, et al30 | 8–12 from different locations | No | Graded into 1 of 3 groups: fibrous, slightly hypervascular, and hypervascular | No | No | DC- MRI is able to distinguish joint effusion from hypervascular pannus and to grade the vascularity of synovitis (only descriptive statistics reported) |
| Tamai, et al21 | 1 sample from each of 3 sites (total no. not defined) | No | 8 histological features assessed SQ (0–3) fibrin exudation, PMN cell infiltration, mononuclear cell infiltration, multiplication of synoviocyte lining layer, villous hypertrophy of synovial surface, proliferation of blood vessels, formation of granulation tissue, and fibrosis | No | No | Rate and degree of signal enhancement with dynamic imaging significantly correlated with histological inflammation (p < 0.05, Mann-Whitney U) |
| Gaffney, et al24 | Not defined | No | SQ histological score 0–3, 3 histological features: PMN infiltration, fibrin, and hyperemia | No | No | Rate of synovial membrane enhancement correlated with histologic features of acute inflammation (r = 0.63, p < 0.01) |
| Gaffney, et al25 | Not defined | No | Not assessed | Endothelial cell marker (Qbend30) assessed by DIA | No | Gd-DTPA–enhanced MRI correlates with histologically determined synovial vascularity (r = 0.55, p < 0.02) |
| Ostergaard, et al22 | 4 biopsy sites (total no. not defined) | SQ macroscopic assessment intra-operatively, 0–3 | 9 features: subsynovial infiltration of PMN leukocytes, subsynovial infiltration of mononuclear leukocytes, surface fibrin deposition, multiplication of the synovial lining, villous hypertrophy of synovial surface, proliferation of blood vessels, perivascular edema, formation of granulation tissue, and fibrosis (SQ score 0–3) | No | No | MRI-determined synovial volumes are correlated with synovial inflammatory activity (r = 0.55, p < 0.001) |
| Ostergaard, et al23 | 4 biopsy sites (total no. not defined) | No | Same as Ostergaard, et al22 | No | No | Early enhancement rate of the total synovial membrane was significantly correlated with histologic grade of synovial inflammatory activity (r = 0.73, p < 10−7) |
| Veale, et al37 | 1 biopsy from each site (total no. not defined) | Hyperemia (0–1), granulation (0–1), or villous hypertrophy (0–2) + overall impression of the synovial inflammation VAS 0–100 mm | Lining layer hyperplasia (SQ 0–3) | T cells [CD3, CD4 (OKT4), CD8], B cells (CD20), macrophages (CD68), and MHC class II.SQ analysis 0–5 | No | The most significant correlations were observed between the MRE at the SPP ROI and arthroscopic VAS for synovitis (r = 0.77, p = 0.003) and between the MRE and immunohistological CD4 score (r = 0.70, p = 0.11) |
| Ostendorf, et al26 | 4–6 biopsies from 2 pts | 6 variables (extent of synovitis, synovial thickening, hyperemia, proliferation, vascularity, fibrosis) + 1 item each for bony and cartilaginous changes (SQ assessment, 0–3) | Synovial hyperplasia, fibrosis, vascularity, lymphocyte and stromal cell infiltration, and fibrin deposition | No | No | Synovial enhancement on MRI correlated with miniarthroscopy findings of hyperemia (p = 0.0038), and vascularity (0.0058). Synovial thickening on miniarthroscopy was significantly associated with synovial proliferation on MRI (p = 0.0063). No formal evaluation of microhistological associations reported |
| Takase, et al27 | Not defined | No | Synovitis score (inflammatory cell infiltrates, synovial lining layer thickness + vascularity), SQ 0–3 | DIA IHC assessment of sublining macrophages (CD68), cell proliferation (Ki67), and neoangiogenesis (CD31) | No | MRI synovitis significantly correlated with total synovitis score (r = 0.48, p < 0.05) and inflammatory cell infiltrates (0.47, p < 0.05) |
| Axelsen, et al28 | 4 (total no. not defined) | No | SQ synovitis score (0–3, 9 histological features) | No | No | Initial rate of enhancement from the quick ROI and the precise ROI revealed high correlations with the grade of histological inflammation (r = 0.70, p < 0.001 and r = 0.74, p < 0.001, respectively) |
| Buch, et al33 | 6 | No | No | T cells (CD3, CD154, CD4), APC (CD80, CD86), B cells (CD20, CD79), synovial fibroblasts (CD55), intracellular adhesion molecules (CD54), macrophages (CD68) and CD11b+ neutrophils, macrophages, and dendritic cells. SQ assessment, 0–4, of all variables in lining and sublining | qRT-PCR (IL-1, IL-6, MMP-1, MMP-3, IFN-γ) | Significant correlation between MRI synovitis scores (initial rate of enhancement and maximum enhancement) and IFN-γ was observed (r = 0.63 and r = 0.79, respectively) |
| Kirkham, et al38 | 3–4 (total no. not defined) | No | SQ analysis of lining layer thickness, vascularity, sublining fibrosis, and cellular infiltrates and patterns (perivascular, diffuse, or focal aggregates) | No | qRT-PCR IFN-γ, TNF-α, IL-16, IL-1β, IL-17, RANKL, and IL-10 | Histologic features had no relationship to damage progression. mRNA levels of IL-1β, TNF-α, IL-17, and IL-10 were predictive of joint damage progression (multivariate regression analysis, r2 = 0.57) |
| Vordenbäumen, et al31 | 6 biopsies | Limited to direction of biopsies | Krenn score39 | Sublining CD68 score | No | Maximum enhancement of the MCP significantly correlated with sublining CD68 staining (r = 0.750, p = 0.02) and synovitis score (r = 0.743, p = 0.02) |
| Vordenbäumen, et al35 | 6 biopsies | No | Krenn score39 | Sublining CD68 score, VEGF, and hypoxia-inducible factor 1α (HIF-1α) | No | VEGF staining correlated with BME (r = 0.676, p = 0.032) and erosion scores (r = 0.695, p = 0.026) of RAMRIS |
| Anandarajah, et al32 | NA | No | Krenn score39 (modified) | No | No | No statistical correlationbetween synovial scores on MRI and synovial hyperplasia on histology |
| Paramarta, et al29 | Assume 6 biopsies | No | No | CD3, CD22, CD68, CD163, vWF | No | Significant association between MRI synovitis and CD68+ sublining macrophage number in RA/SpA (r = 0.686, p = 0.001) |
| Kennedy, et al34 | NA | Synovitis and vascularity (VAS 0–100 mm) | Not reported | CD3, CD68, CD4, CD8, CD20, CD19, and factor VIII/αSMA | No | Change in MRI score after TNF blocking therapy was significantly associated with changes in macroscopic synovitis (p = 0.056), sublining CD68 (p < 0.002), and sublining CD4 cell number (p = 0.032) |
| Maijer, et al36 | 6 biopsies | No | No | vWF | No | Demonstrates that DCE-MRI pharm. variables differ between different diagnostic categories and correlate with local (synovial vWF) and systemic markers of disease activity |
MRI: magnetic resonance imaging; IHC: immunohistochemical; NA: not available; SQ: semiquantitative; VAS: visual analog scale; PMN: polymorphonuclear; DIA: digital image analysis; VEGF: vascular endothelial growth factor; vWF: von Willebrand factor; qRT-PCR: quantitative real-time PCR; IL-1: interleukin 1; MMP: matrix metalloprotease; IFN-γ: interferon-γ; TNF-α: tumor necrosis factor-α; RANKL: receptor activator of nuclear factor-κB ligand; Gd: gadolinium; SPP: suprapatellar pouch; pharm.: pharmacokinetic; ROI: region of interest; MCP: metacarpophalangeal; BME: bone marrow edema; RAMRIS: Rheumatoid Arthritis Magnetic Resonance Imaging Scoring; RA: rheumatoid arthritis; SpA: spondyloarthritis; DCE: dynamic contrast-enhanced; APC: antigen-presenting cell; HIF: hypoxia-inducible factor; MRE: maximal rate of enhancement.