Synovial lining, endothelial and inflammatory mononuclear cell proliferation in synovial membranes in psoriatic and reactive arthritis: a comparative quantitative morphometric study

Abstract

The extent of synovial cell proliferation in situ and its relationship to the destructive potential of rheumatoid arthritis (RA) is a matter of continuing debate. Notably, the situation has not been elucidated in other inflammatory arthritides [i.e. reactive (ReA) and psoriatic (PsA)], which, although they share some histopathological similarities with RA, develop different patterns of joint involvement. In order to estimate the proliferation of synovial cells in situ in PsA and ReA, and to compare this with RA and with 'non-inflammatory' joint lesions, we have utilized immunostaining of the Ki-67 antigen complemented with Ki-67/CD68 or Ki-67/leucocyte common antigen (LCA, clones 2B11 and PD7/26) double stainings to assess the extent of mononuclear inflammatory cell proliferation. Synovial samples analysed were from 33 patients: RA (n = 8), PsA (n = 13), ReA (n = 6) and six 'non-inflammatory controls' (degenerative or traumatic joint lesions). Thickening of the synovial lining (in particular in RA) and perivascular accumulations of mononuclear inflammatory cells, predominantly lymphocytes, were characteristic features in all synovitides. In contrast to the thickened avascular synovial lining in RA, in 5/13 cases with PsA, blood vessels were observed in the lining. The percentage of lining cells expressing Ki-67 antigen was higher in RA (median = 4.7, interquartile range [Q3-Q1] = 3.9, mean [95% CI] = 3.5 [1.7-5.2], P = 0.0063), PsA (median = 1.2, [Q3-Q1] = 1.9, mean [95% CI] = 1.6 [0.7-2.5], P = 0.007) and ReA (median = 1.4, [Q3-Q1] = 2.3, mean [95% CI] = 1.6 [0.1-3.1], P = 0.0235) than in controls (median = 0.1, [Q3-Q1] = 0.45, mean [95% CI] = 0.2 [0.07-0.5]). In this respect, the differences between different forms of the inflammatory arthritides were not statistically significant (P > 0.05). In RA, PsA and ReA, the percentage of labelled cells in the inflammatory mononuclear cell-rich areas was higher than in controls. The percentage of proliferating endothelial cells was also significantly higher in RA, PsA and ReA than in controls. However, in RA, endothelial expression of Ki-67 antigen was often seen in small blood vessels, whereas in PsA, Ki-67 antigen was preferably expressed in the medium to large blood vessels. Synovial lining cells of the monocyte/macrophage lineage (type A synoviocytes), but not stromal monocytes, demonstrated modest proliferation in situ. These results indicate that although proliferation of synovial lining fibroblasts is a prominent feature in RA, the extents to which this, or in situ proliferation of lymphocytes, contribute to the histopathology of PsA, ReA and RA are comparable. Vascular involvement is suggested by the proliferation of endothelial cells in RA, PsA and ReA in an overlapping manner, but, based on topological differences, such a response may represent diverse pathological features, such as angiogenesis, vascular enlargement and reparative responses to injury.