Article Text

Download PDFPDF

Immunohistochemistry of normal synovium
  1. M D Smith1
  1. 1Rheumatology Research Unit, Repatriation General Hospital, South Australia 5041, Australia
  1. Correspondence to:
    Dr M D Smith
  1. J A Singh2,
  2. H R Schumacher2
  1. 2Minneapolis VA Medical Center, Minneapolis, USA

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

A recent paper published in the Annals of the Rheumatic Diseases by Singh et al claimed to be the first report of immunohistochemical features of knee synovium in such a large number of healthy normal subjects.1 This statement is incorrect and I am surprised that the authors of this paper appear to be unaware of our paper published more than 12 months previously in the same journal.2

We studied 20 normal patients attending a sports medicine clinic with unexplained knee pain who had no evidence of any form of arthritis on history, examination, and laboratory tests and had normal x ray findings and normal knee arthroscopy. Although Singh et al may argue about the source of our “normal” subjects, it is noted that several of their subjects were not entirely normal either. We are, at least, in a better position to state that our patients were as close as possible to normal than the study of Singh et al, who relied on normal x ray examinations to exclude joint pathology (including early osteoarthritis) and included patients with either a positive rheumatoid factor or a raised erythrocyte sedimentation rate (ESR), neither of which was included in our patient group.

In the two studies the patient group and the variables measured in the synovial membrane were similar, and the conclusions of both studies were also similar. Like the study of Singh et al, we found quite a variability in the architecture of the normal synovial membrane, particularly in relation to the thickness of the lining layer and the subintimal cell infiltrate, but we also measured cytokine production, cell adhesion molecule expression, and mediators of osteoclast formation, which were not included in the study of Singh et al. We did find evidence of B cells and occasional plasma cells in the normal synovial membranes, unlike Singh et al, but these cell populations were quite sparse.

It is, however, very surprising that neither the journal reviewers nor the authors of this paper appeared to be aware of a similar paper published in the same journal before this paper was accepted for publication, and we do dispute the statement of Singh et al that they have published the “first report of immunohistochemical features of knee synovium in such a large number of healthy normal subjects”.


Authors’ reply

We thank Dr Smith for alerting readers and us to the existence of his excellent paper that came out after our study was done and during the preparation of our manuscript.

Our study described asymptomatic subjects,1 a somewhat different group from that in Dr Smith’s study2—namely, patients with unexplained knee pain with clinically and arthroscopically normal knee joints. It is fascinating that in his study arthroscopy could not detect moderate inflammation seen in some biopsies. Thus, some synovitis can be present even when examinations and gross arthroscopy are normal. Future studies of patients like ours and those of Smith et al would benefit from longer follow up. What was the final cause of the knee pain that led to the need for arthroscopy? Will the cellular pattern be consistent, resolving, or progressive in some patients? Do the infiltrates seen in some “normal subjects” have any prognostic significance?

It appears that our two studies1,2 with many similar findings reinforce the point, also noted before by Lindblad and Hedfors,3 that there may be more variability in synovial histology and immunohistochemistry than many may have appreciated.