Article Text

Download PDFPDF
Pitfalls for detecting interleukin-33 by ELISA in the serum of patients with primary Sjögren syndrome: comparison of different kits
  1. Elodie Rivière1,2,
  2. Bineta Ly2,
  3. Saida Boudaoud2,
  4. Houria Chavez2,
  5. Gaetane Nocturne1,2,
  6. Philippe Chanson3,
  7. Corinne Miceli Richard1,2,
  8. Xavier Mariette1,2
  1. 1 Rheumatology Department, Université Paris-Sud, Assistance Publique-Hôpitaux de Paris, Le Kremlin-Bicêtre, France
  2. 2 INSERM U1184, Université Paris-Sud, Le Kremlin-Bicêtre, France
  3. 3 Endocrinology Department, Université Paris-Sud, Assistance Publique-Hôpitaux de Paris, Le Kremlin-Bicêtre, France
  1. Correspondence to Professor Xavier Mariette, Hôpitaux Universitaires Paris-Sud, Université Paris-Sud, 78 rue du Général Leclerc, Le Kremlin Bicêtre 94270, France; xavier.mariette{at}

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.

The recent interest for interleukin-33 (IL-33) in primary Sjögren syndrome (pSS) pathophysiology is due to three studies,1–3 which have shown an increase in IL-33 serum levels in patients with pSS compared with controls. IL-33 is a member of the IL-1 family which is released in case of cellular damage and acts as an alarmin; however, its role in pSS physiopathology remains unclear.

In the present study, we aimed to replicate the association described between IL-33 serum level and pSS diagnosis. For that purpose, we used Quantikine and Duoset ELISA (R&D Systems, Minnesota, USA) and Luminex (Milliplex HTh17, Merck, Darmstadt, Germany). The Quantikine and Luminex kits are validated for serum samples, while the Duoset is only validated for use with culture supernates (data from R&D). IL-33 serum levels were assessed in patients with pSS from the French multicentre prospective ASSESS cohort (Assessment of Systemic Signs and Evolution in Sjögren Syndrome, principal investigator: X M) and in two types of controls: subjects with sicca, asthenia, polyalgia syndrome (SAPS) without anti-SSA or anti-SSB antibodies and no focal lymphocytes infiltrates in minor salivary gland biopsies, and healthy controls from the Variété cohort (principal investigator: PC). Duoset and Quantikine ELISA and Luminex were performed according to manufacturer's instructions, and sera were not diluted. Results were expressed in pg/mL as mean±SD.

First, we tested sera with IL-33 Quantikine ELISA. No significant difference was observed for IL-33 serum levels between pSS (n=59) (0.7498±3.518 pg/mL) and the two control groups SAPS (n=46): 0.1738±1.179 pg/mL, Variété (n=16): 0.9213±3.685 pg/mL (figure 1A). Surprisingly, most of the sera tested (116/121) were below the minimal detection limit of the kit. The results of Luminex confirmed the absence of IL-33 serum level augmentation in patients with pSS (n=32) 24.87±132pg/mL compared with controls (Variété, n=7) 45.14±69.68 pg/mL (figure 1B). This was not expected given prior published data. However, the three published studies used the Duoset version of the ELISA.

Figure 1

(A) Serum levels of interleukin-33 (IL-33) in patients with primary Sjögren syndrome (pSS) (from the ASSESS cohort) and controls (from sicca, asthenia, polyalgia syndrome (SAPS) and Variété cohort). Dosage performed with Quantikine ELISA IL-33 kit (R&D Systems). ns: non-significant. Analyses were performed with the GraphPad software. (B) Serum levels of IL-33 in patients with pSS from the ASSESS cohort and Variété controls. Dosage performed with Luminex (Merck).

Therefore, to understand this discrepancy, we have compared the results of pSS and controls samples tested with the Quantikine and the Duoset ELISA (n=12). We have shown a total absence of relation between results of both kits: among nine patients with IL-33 serum levels below the minimal detection limit with the Quantikine, results of the Duoset varied from 0 to 1500 pg/mL (figure 2). Moreover, we have now performed other comparisons between both kits in a larger number of patients with rheumatoid arthritis (n=129) and controls (n=46) and confirmed the absence of correlation (r=0.02311, p=0.9758) (data not shown). Spike recovery tests consist of adding a known amount of the cytokine into a negative serum sample, to detect if the addition of the serum sample does not change the expected value. Using this test, we found that sera components were likely to interfere with the reagents used in the Duoset kit. Conversely, there was no interference between negative sera and the Quantikine kit.

Figure 2

Comparison between serum interleukin-33 (IL-33) levels determined with the Quantikine ELISA IL-33 kit and with the Duoset ELISA IL-33 kit (R&D) in primary Sjögren syndrome (pSS) and controls sera (n=12).

In conclusion, Duoset ELISA IL-33 kit, which has not been validated for human sera, should not be used for IL-33 dosage in serum. Results already obtained with the Duoset kit concerning IL-33 serum levels should be interpreted with extreme caution. Using the Quantikine ELISA validated for sera, we did not find any increase of IL-33 serum level in patients with pSS.


We thank Eric Lefèvre, PhD, Bio-Techne, UK for providing Duoset kits and for helping with the interpretation of results.


View Abstract


  • Contributors ER designed and performed experiments, analysed data and wrote the manuscript. BL performed experiments and analysed data. SB, GN and CMR designed experiments and analysed data. PC provided serum samples for Variété controls. XM designed experiments, analysed data and revised the manuscript.

  • Funding We thank the French Society of Rheumatology which provided financial support.

  • Competing interests None declared.

  • Ethics approval Local Ethic Committee of Hôpital Bichat and the ‘Comission Nationale informatique et Liberté’ in 2006.

  • Provenance and peer review Not commissioned; externally peer reviewed.

  • Data sharing statement All the authors in this current work declare that there are no unpublished data and that all data have been presented in the current manuscript.