Article Text


Diagnostic value of anti-cyclic citrullinated peptide antibodies to detect rheumatoid arthritis in patients with Sjögren’s syndrome
  1. C van Noord,
  2. H Hooijkaas,
  3. B C M Dufour-van den Goorbergh,
  4. P M van Hagen,
  5. P L A van Daele,
  6. J P van de Merwe
  1. Departments of Immunology and Internal Medicine, Erasmus MC, University Medical Centre Rotterdam, The Netherlands
  1. Correspondence to:
    MsC van Noord
    Erasmus MC, University Medical Center Rotterdam, Department of Immunology, Dr Molewaterplein 50, 3015 GE Rotterdam, The Netherlands;

Statistics from

Sjögren’s syndrome (SS), prevalence 3–4%,1 is a chronic autoimmune disorder characteristically affecting the salivary and lachrymal glands. Rheumatoid arthritis (RA), prevalence 1–4%, is a chronic inflammatory autoimmune disease.2

The diagnosis of RA relies mainly on clinical manifestations and serological markers such as rheumatoid factors (RF). The sensitivity of RF in RA is 75% and the specificity 74%.3 Furthermore, RF is positive in 40–70% of patients with primary SS.4 Many patients with primary SS and chronic polyarthritis consequently have RF without ever developing RA. An enzyme linked immunosorbent assay (ELISA) test has been developed that recognises a cyclic variant of a citrullinated peptide (CCP).2 The sensitivity of the first generation anti-CCP test in RA ranges from 41 to 68%,5,6 the sensitivity of the second generation is 82%.7 The specificity, however, is 96–98%.3,5–7

We analysed data from 164 patients who were diagnosed as SS according to the revised version of the European criteria.8 These criteria allow a diagnosis of SS if at least four items out of six or three objective items are present. Unfortunately, no single laboratory test is sufficiently reliable to confirm a clinical diagnosis of SS.9 Therefore, a second group was assembled with patients in whom three items were present and in whom no other disease could explain the sicca symptoms. This group is further referred to as Sjögren’s-like syndrome.

The medical records from all patients were further investigated for RA, according to the 1987 revised criteria.10 RF and anti-CCP antibodies were determined in the same serum samples using the ELISA anti-CCP mark 2 (second generation) kits from Immunoscan RA, Euro-Diagnostica AB (Arnhem, Netherlands) and the IgM RF ELISA test. All the data were analysed using the SPSS/PC software, version 11.0.

Table 1 shows that both groups were similar. Furthermore, it shows that anti-CCP has a high specificity (98.8%), in contrast with the low specificity of RF (60.6%).

Table 1

 Characteristics of all patients

The diagnostic value of the RF test in patients with SS is questionable because of its low specificity (60.6%) in such patients. In this study we found a specificity of 98.8% for anti-CCP in the SS population for RA. The major strength of these data is to emphasise the fact that anti-CCP is not present in patients with primary SS who do not have RA, in contrast with the high prevalence of RF in patients with primary SS.

Figure 1 shows an overview of patients who have RA. Patients A and D have SS and RA, patients B and C have Sjögren’s-like syndrome and RA while patients E and F do not have RA (RF was negative in both patients), but have a borderline anti-CCP. They could develop RA at a later stage. Patients A, B, and C have a positive anti-CCP.

Figure 1

 Timescale of the four patients with RA with the relevant anti-CCP results. Patient A (SS) 1991 RF: 50 IU/ml, 2002 RF: 200 IU/ml; patient B (SS-like) 2003 RF: 1600 IU/ml; patient C (SS-like) 2002 RF: 50 IU/ml, 2003 RF: 50 IU/ml; patient D (SS) 1994 RF: 50 IU/ml, 2002 RF: 12 IU/ml. The evaluation criteria for anti-CCP are as follows: ⩽25 U/ml are defined as negative, >50 U/ml are positive. Samples with >25 U/ml and <50 U/ml were considered as borderline. The evaluation criteria for RF are as following: negative (<12 IU/ml) and positive (⩾12 IU/ml).

In this study only four patients were diagnosed with RA. This limits the accuracy of the sensitivity, but is excellent for calculating the specificity. The rationale for this study is the fact that it is important to select only those patients with RA from a group in which most have a positive RF test, with the help of a specific diagnostic test such as the anti-CCP test. When a cut off value of 100 U/ml for anti-CCP is used, the specificity of anti-CCP for RA is 100%.

We conclude that the RF test for the diagnosis of RA in patients with SS has no value because about 40% of patients with SS have positive RF tests (94% of these tests are false positive). The anti-CCP test, on the other hand, has a high specificity for RA.


We thank Jane Meijlink for her help with the English translation.


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.