Article Text

Download PDFPDF

Sjögren's syndrome criteria
  1. M M Zandbelt1,
  2. F H J van den Hoogen1
  1. 1Department of Rheumatology, University Medical Centre Nijmegen, The Netherlands
  1. Correspondence to
    Dr M M Zandbelt;
  1. R Manthorpe2
  1. 2Sjögren's Syndrome Research Centre, Department of Rheumatology, Malmö University Hospital, SE-205 02 Malmö, Sweden

    Statistics from

    In the June issue of the Annals Manthorpe comments on the recently proposed US-European classification criteria for Sjögren's syndrome (SS).1 We would like to deal with some of the issues he raises, and add some comments.

    Now that the classification criteria have evolved from rather subjectively biased ones to more objective assessments, it is surprising that the two most disease-specific objective parameters currently available for SS are subject to considerable criticism. Of course, when serological and histological items are emphasised in the new SS classification criteria, their individual disease sensitivity and specificity should always be kept in mind.

    In fact, all six items that are included in the classification criteria may be subject to discussion. For example, the Schirmer-I test, and unstimulated whole salivary flow test have been criticised in a number of papers,2–5 but these items are recommended in Manthorpe's paper.

    Manthorpe expresses his concerns about the accuracy of sublabial salivary gland biopsies (SLGBs), referring to one paper in which a change of diagnosis of >50% is reported after a second examination of the SLGBs. However, the authors themselves report that not using the focus scoring system was probably the most important reason for the change of diagnosis on the second examination. They did not conclude that the focus score itself—which is mandatory to fulfil item VI—changed dramatically upon re-examination of the specimens!

    Other ways of bypassing interobserver variability are also available—for example, measuring two parameters instead of one (for example IgA% and focus score) provides a synergistic value for the accuracy of diagnosis,6 and, moreover, computer aided scoring methods may provide non-observer dependent data. For measuring the IgA% reliable and reproducible objective data from the biopsies are obtained by combining microscope, computer, and calibrated software. These biopsies show what is going on in the target organs of this disease and may provide early diagnostic markers; one should not put them aside too easily.

    Manthorpe also criticises the SS classification criteria for the interdependent relation between anti-Ro/anti-La antibodies (item IV) and the focus score (item VI). They are certainly associated with each other, but why is that a problem ? The worldwide accepted American Rheumatism Association criteria for rheumatoid arthritis also contain interdependent items—for example, positive rheumatoid factor serology is generally considered as strongly associated with radiological joint damage. Interdependency can also be found in the American College of Rheumatology classification criteria for systemic lupus erythematosus (presence of antinuclear antibodies is a distinct item from presence of anti-dsDNA or anti-Sm, items 11 and 10 respectively). Furthermore, it appears inconsistent that Manthorpe recommends including the patient's smoking habits in the SS classification criteria. This would also introduce an interdependent item.

    The dependency does not equal a one-on-one relation—that is, seronegative patients may have a positive focus score and vice versa. In particular, because numerous reports have shown that the focus score alone can be false positive or false negative,7–10 the presence of anti-Ro/anti-La antibodies, which are still the most disease-specific and sensitive parameters available, has additional value for the accuracy of diagnosis. Finally, it has yet to be proved that the suggested new antibodies (anti-fodrin, anti-muscarin) are more sensitive and disease-specific than the existing classic anti-Ro and anti-La antibodies. Therefore it is too early to include such items in classification criteria.

    While our knowledge of Sjögren's syndrome increases, classification criteria may develop in a way that enhances early diagnosis of possibly reversible target organ damage. Therefore not the end stage symptoms and signs (items I–III and V) but rather the early target organ histological signs and serological signs are likely to retain their place in the classification criteria. Therefore, in our view the US-European consensus group is right to emphasise items IV and VI, which should not be neglected until better alternatives have been introduced.


    Author's reply

    Zandbelt and van den Hoogen raise and discuss some important issues which I put forward in the June Leader of the Annals concerning Sjögren's syndrome (SS) criteria published by a consensus group consisting of European and North American SS experts. The subtitle read: “American-European (US-Eur) and Japanese Groups' criteria compared and contrasted.” Zandbelt's and van den Hoogen's points are well taken, although the issues put forward are not new. I agree with them that final diagnostic criteria will arrive on the day when we know the aetiopathogenesis. Until that happens we trust (and are stuck with) classification criteria that are primarily set up as research tools but nevertheless find their way into daily clinical practice. I am of the opinion that it is best to have as few preliminary classification criteria as possible. When coming up with new proposals these should include changes that are up to date in all aspects, otherwise other proposals will arise too soon. It is here, among other things, that I am disappointed by the consensus group's latest proposal.

    Most SS specialists agree that it is difficult to diagnose SS without close collaboration between clinical specialists within ophthalmology, oral medicine/oral surgery, and rheumatology. The US-Eur proposal is written by 13 authors but not a single person is an ophthalmologist! We have known for some years that patients with genuine SS do not complain of dry eyes because the cornea—although heavily innervated—lacks nerves that register dryness. This makes item I in the proposed criteria set invalid and should have been changed.

    As far as smoking is concerned, it seems of great importance that clinicians know such details just as they know the medical history. When did smoking start and stop? What was the weekly consumption of cigarettes? Patients, who are present or past smokers, and who have at least two abnormal objective test results from both the main affected exocrine organs, lachrymal and salivary glands, very often lack circulating anti-SSA/B autoantibodies and simultaneously have a focus score ≤1.

    As the greatest percentage of the world population is present or past smokers with a consumption of >21 cigarettes/week some will have eye and oral symptoms similar to patients with SS (item I updated plus item III), and I find it difficult to accept that such patients would not be diagnosed as having SS if the US-Eur criteria were followed. Given that the patient in question is a present or past smoker should lead to the consequence that a focus score ≤1 and/or absence of anti-SSA/B autoantibodies (item IV and VI) cannot be trusted and consequently should be disregarded. Besides personal/family consequences it might have great social effects in some countries. In Sweden, for example, patients might get their dental repair bill subsidised by the State if they have SS diagnosed according to the Copenhagen criteria and, in addition, have abnormal unstimulated and stimulated whole sialometry, measured by 15 and 5 minutes, respectively.

    I agree with Zandbelt and van den Hoogen that we do not have specific SS autoantibodies and neither do I think that the last autoantibody has been found. Classification criteria should not, therefore, concentrate solely upon the SSA/B autoantibodies but be open to newer ones as well. As mentioned in the leader the newly discovered BAFF (B cell activating factor from the tumour necrosis factor family) seems promising.

    The fact that Japanese SS specialists simultaneously present their new classification criteria (also termed Japanese III), which look rather different from the US-Eur consensus criteria, might be considered very disturbing and disappointing for clinicians. However, they do look more acceptable, are based upon the results of a greater number of patients, and focus more on objective assessments—as asked for by Zandbelt and van den Hoogen. However, Zandbelt and van den Hoogen forget that the original set of SS criteria, the Copenhagen criteria of 1975, were based purely on objective data. The history of the various classification criteria for SS from that date can simply be represented by a nearly closed circle (see fig 1 in the June leader).1

    At the VIIIth International SS symposium held in mid-May 2002 in Kanasawa, Japan, both the new criteria were presented and discussed. The proposal put forward by the president of the symposium, Professor Susumu Sugai, that an international group of SS researchers should be inaugurated with representatives from Europe, America, China, and Japan was very much applauded. I hope that this international group of SS experts will not repeat the error from the systemic lupus erythematosus (SLE) criteria,2 where the specificity of the proposed SLE criteria were tested against only two cases of SS. My qualified guess would be that the final proposal from this international group will not look like the newly presented US-Eur criteria. These criteria would seem to have a rather short timespan and therefore it may not pay to introduce them immediately as classification criteria for use in the clinic.


    View Abstract

    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.