Article Text

Download PDFPDF

Short report
Anti-dsDNA IgE: a potential non-invasive test for prediction of lupus nephritis relapse
  1. Marie Himbert1,
  2. Noémie Jourde-Chiche2,3,
  3. Léa Chapart4,
  4. Nicolas Charles4,
  5. Karine Baumstarck5 and
  6. Eric Daugas1,4
  1. 1Department of Nephrology, Hôpital Bichat Claude-Bernard, Paris, France
  2. 2C2VN, INSERM, INRAE, Aix-Marseille Universite, Marseille, France
  3. 3Centre de Néphrologie et Transplantation Rénale, AP-HM, CHU de la Conception, Marseille, France
  4. 4Centre de Recherche sur l'inflammation, INSERM UMR1149, CNRS EMR8252, Université Paris Cité, Paris, France
  5. 5Laboratoire de Santé Publique CERESS, Aix-Marseille Université, Marseille, France
  1. Correspondence to Dr Eric Daugas; eric.daugas{at}


Objectives Discontinuation or continuation of maintenance immunosuppressive therapy (MIST) after a severe lupus nephritis (LN) requires measuring the risk of relapse but reliable clinical and biological markers are lacking. The WIN-IgE study assesses the value of serum anti-dsDNA IgE autoantibodies as a biomarker for the prediction of relapse in severe LN.

Methods WIN-IgE is an ancillary study of the WIN-Lupus study (NCT01284725), a prospective controlled clinical trial which evaluated the discontinuation of MIST after 2–3 years in class III or IV±V LN with active lesions. WIN-IgE included all patients with available serum collected at randomisation for continuation or discontinuation of MIST. In these sera, anti-dsDNA antibodies, IgE and IgG, were quantified by ELISA and compared between patients who experienced LN relapse and those who did not during the 24 months of follow-up.

Results 52 patients were included, 25 in the MIST continuation group and 27 in the MIST discontinuation group, 12 experienced a biopsy-proven relapse of LN. Initial anti-dsDNA IgE antibodies levels were higher in patients with subsequent LN relapse. Anti-dsDNA IgG was not associated with relapse. Survival without LN relapse was lower in patients with anti-dsDNA IgE levels above vs below a threshold of 1.9 arbitrary units (p=0.019), particularly in the subgroup of patients randomised to discontinue MIST (p=0.002). In all patients, anti-dsDNA IgE above 1.9 arbitrary units had a positive predictive value of 0.8 for severe LN relapse.

Conclusions These results suggest blood anti-dsDNA IgE as a non-invasive predictive marker of LN relapse.

  • Systemic Lupus Erythematosus
  • Lupus Nephritis
  • Risk Factors

This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See:

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.

What is already known on this topic

  • After severe lupus nephritis (LN), there are no reliable non-invasive markers of risk of relapse to guide the continuation or discontinuation of maintenance immunosuppressive therapy (MIST).

What this study adds

  • High levels of anti-dsDNA IgE were associated with a higher risk of relapse of severe LN, particularly in patients who had discontinued MIST.

How this study might affect research, practice or policy

  • Blood anti-dsDNA IgE is a candidate biomarker for the non-invasive assessment of the risk of severe LN relapse and for personalising the continuation or discontinuation of MIST.


Proliferative lupus nephritis (LN) (class III or IV LN in the ISN/RPS 2003 classification, with active lesions) can lead to chronic kidney disease or end-stage kidney disease. Their treatment consists of immunosuppressive therapy based on an induction phase to achieve remission, followed by a maintenance phase to prevent relapses. The optimal duration of maintenance immunosuppressive therapy (MIST) is still debated, and recent recommendations1 2 are based on expert opinion.

The WIN-Lupus study (NCT01284725),3 a randomised controlled trial, compared discontinuing MIST after 2–3 years with continuing MIST for another 2 years in patients with class III or IV LN in stable remission on hydroxychloroquine and MIST with azathioprine or mycophenolate mofetil. The hypothesis, which was the non-inferiority of MIST discontinuation in terms of LN relapse after 2 years, was not demonstrated. On the contrary, MIST discontinuation was associated with an increased risk of severe systemic lupus erythematosus (SLE) flare (renal or extra-renal) at 2 years.3 However, the majority of patients (about 70%) who discontinued MIST did not experience a relapse of LN or a severe flare. The current challenge is therefore to assess the individual risk of LN relapse in order to tailor MIST. Some clinical risk factors for LN relapse were identified in WIN-Lupus, such as higher baseline urinary protein to creatinine ratio or low C3, but anti-dsDNA IgG antibodies were not associated with relapses, as shown in previous studies.4–7 Control kidney biopsy, to assess histological remission in addition to clinical remission, has recently been shown to predict LN relapse and may be used to guide MIST weaning.8 However, kidney biopsy is an invasive procedure and safer biomarkers would be welcome. The involvement of autoreactive anti-dsDNA IgE antibodies has been highlighted in the establishment of an amplification loop of SLE activity, particularly in LN, via basophil and plasmacytoid dendritic cells (pDCs) activation.9

Here, we aimed to assess the value of serum anti-dsDNA IgE autoantibodies as a biomarker of the risk of relapse of proliferative LN in patients from the WIN-Lupus trial.


We used the serum bio-collection from WIN-Lupus, which comprised sera drawn at randomisation (M0) for MIST continuation or discontinuation. All patients from the per-protocol population with available sera at M0 were included in this ancillary study, the WIN-IgE study. Anti-dsDNA antibodies, of both IgE and IgG isotypes, were quantified by indirect ELISA as previously described10 and detailed in online supplemental methods. Anti-dsDNA antibody levels were compared between patients who experienced or did not experience a biopsy proven relapse of LN during the 24-month follow-up from M0 (the time of the anti-dsDNA antibodies test) to M24. The optimal discrimination threshold of anti-dsDNA IgE was identified by the Youden index, defined by the maximum value of sensitivity+specificity−1, for anti-dsDNA Ig values above the threshold of positivity (set at 1). Kaplan-Meier analyses were performed to compare relapse-free survival of patients according to this threshold as detailed in online supplemental methods.


Among the 84 patients of the per-protocol population of WIN-Lupus, 52 patients had a serum sample available at M0 and were included in the present study: 25 were in the MIST continuation group, and 27 in the MIST discontinuation group (online supplemental figure S1). Of these 52 patients, 12 (23%) experienced a biopsy-proven relapse of LN, and 40 did not experience a LN relapse until M24. Most patients were women (90%), most patients were Caucasian (73%), mean age was 39 years, and most patients (79%) had presented a first flare of LN (they were not LN relapsers) before their inclusion in WIN-Lupus (table 1 and online supplemental table S1).

Table 1

Main characteristics of patients in the WIN-IgE study according to their LN relapse status during follow-up and according to their randomisation group in the WIN-Lupus study

Anti-dsDNA IgE and IgG levels did not differ between randomisation groups at M0 (online supplemental figure S2).

Anti-dsDNA IgE antibodies were predictive of LN relapse: their positivity at M0 (p=0.04) and their level at M0 (p=0.025) (online supplemental figure S3) were both associated with LN relapse between M0 and M24. The association between anti-dsDNA IgE levels and LN relapse was particularly striking in patients randomised in the MIST discontinuation group (online supplemental figure S4). In contrast, anti-dsDNA IgG was not associated with subsequent relapse (online supplemental figure S5).

A discriminating threshold level of 1.9 arbitrary units—corresponding to 1.9 times the positivity threshold—was determined for anti-dsDNA IgE to predict a severe LN relapse (online supplemental figure S6). Its specificity for severe LN relapse was 0.9, while its sensitivity was 0.4. Its positive predictive value for LN relapse was 0.8 and its negative predictive value was 0.6. Survival without LN relapse differed significantly (p=0.019) between patients with baseline anti-dsDNA IgE levels below and above the threshold of 1.9 in the overall patient population, with relapse-free survival being lower in patients with levels above 1.9. However, this difference was only observed in the MIST discontinuation group (p=0.002, figure 1) and not in the MIST continuation group (not significant, figure 2).

Figure 1

Survival without lupus nephritis (LN) relapse, in patients randomised in the maintenance immunosuppressive therapy (MIST) discontinuation group. Survival without relapse of lupus nephritis, in patients randomised in the MIST discontinuation group, according to their baseline (M0) anti-dsDNA IgE levels above vs below the threshold of 1.9.

Figure 2

Survival without lupus nephritis (LN) relapse in patients randomised in the maintenance immunosuppressive therapy (MIST) continuation group. Survival without relapse of lupus nephritis, in patients randomised in the MIST continuation group, according to their baseline (M0) anti-dsDNA IgE levels above vs below the threshold of 1.9.


These results highlight the potential value of less classical anti-dsDNA antibodies, of the IgE isotype, in predicting LN relapse, particularly in patients who discontinue MIST. Indeed, anti-dsDNA IgE positivity and baseline level both predicted LN relapse in patients from the WIN-Lupus randomised controlled trial.

DNA-anti-DNA IgE complexes are equivalent to DNA-anti-DNA IgG complexes in their ability to induce TLR9-dependent interferon alpha (IFNα) production by pDCs and the presence of both isotypes in these complexes leads to synergistic effects.11 Despite the lower abundance of IgE as compared with IgG in the serum, an increase in the anti-dsDNA IgE/IgG ratio in circulating immune complexes may trigger a stronger production of IFNα by pDCs. Furthermore, anti-dsDNA IgE can activate an amplification loop of autoantibody production involving basophils,9 12 a property not established for anti-dsDNA IgG. In this sense, the involvement of anti-dsDNA IgE not only complements the involvement of anti-dsDNA IgG in the pathophysiology of lupus flares, but also participate in the most severe forms, particularly LN.9 10 Complementarily, we show here that anti-dsDNA IgE antibodies are associated with the risk of LN relapse in patients in remission. Note that anti-dsDNA IgE then appears to be more consistently associated with lupus activity than total IgE, which have been proposed to have a positive association with activity12 13 or conversely, a protective role.14

The 60% negative predictive value of an anti-dsDNA IgE test below 1.9 arbitrary units is low and of limited clinical interest. Conversely, the 80% positive predictive value of anti-dsDNA IgE above 1.9 units, established in a population where the overall risk of LN relapse is 23%, may be helpful in distinguishing patients at risk of relapse when discussing continuation or discontinuation of MIST, in addition to other risk factors such as extrarenal lupus activity, higher baseline urinary protein to creatinine ratio or low C3 level.

In contrast, this study finds a poor value of anti-dsDNA IgG in predicting LN relapse, as previously discussed in the literature.15 We used an ELISA for their detection that may have a lesser ability for detection of high-avidity anti-dsDNA IgG than the reference test, the Farr assay. Then, ELISA-detected anti-dsDNA IgG in the present study may be less reliable for predicting clinical activity of lupus due to their possible lower avidity for dsDNA. Nevertheless, we used similar ELISAs for IgG and IgE anti-dsDNA detection, which allows comparison and the conclusion that IgE anti-dsDNA is probably more efficient than IgG anti-dsDNA for predicting LN relapse, at least in this study.

The small sample size of the WIN-IgE study—52 patients—and the low number of events—12 relapses (five in the continuation group and seven in the discontinuation group)—are limitations. However, there are few studies with well standardised patients followed prospectively, as here, allowing the investigation of MIST discontinuation. WIN-IgE should then be analysed as a preliminary study. Its results need to be confirmed in larger and multi-ethnic cohorts, and the value of anti-dsDNA IgE put into perspective with that of other risk factors for LN relapse, and possibly with that of repeat kidney biopsy. WIN-Lupus allowed us to identify some risk factors for relapse among usual clinical and biological characteristics of patients at inclusion: higher urinary protein/creatinine ratio at baseline, low C3, higher SLE Disease Activity Index, lower leucocyte, lymphocyte, basophils counts, lower serum albumin, lower haemoglobin—which could be directly or indirectly associated with lupus activity—as well as higher estimated glomerular filtration rate, lower eosinophil counts and antiphospholipid syndrome, were all associated with further LN relapse. However, to support their use in clinical practice, these factors, as well as anti-dsDNA IgE, should be confirmed in replication studies and, more importantly, compared and hierarchically ranked by multifactorial analyses, which was not possible due to the small number of patients in both WIN-Lupus and WIN-IgE.

In conclusion, this ancillary study of the WIN-Lupus trial suggests the specific value of anti-dsDNA IgE autoantibodies as a non-invasive predictive marker of LN relapse, especially in patients for whom MIST is discontinued. The classical anti-dsDNA IgG auto-antibodies were less useful in this regard. Anti-dsDNA IgE autoantibodies could be used in patients, in addition to other factors such as lupus clinical activity, C3 level, residual proteinuria, residual active histological lesions on repeated kidney biopsies, to assess the risk of LN relapse after 2–3 years of MIST and to guide subsequent treatment management, as part of a therapeutic personalisation approach.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants. The trial was conducted in accordance with the principles of the Declaration of Helsinki and the International Conference on Harmonization Guidelines for Good Clinical Practice, addendum E6. It was approved on 30 September 2010 by the Comité de Protection des Personnes Sud-Méditerranée I (ID-RCB 2010-022859-30). All patients gave their written informed consent before any study-related procedure.


Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Contributors ED, NJ-C and NC designed the study. MH and LC performed laboratory analysis. KB and MH performed statistical analyses. All authors analysed the results. MH and ED wrote the manuscript, all authors reviewed and approved the manuscript. ED is guarantor.

  • Funding The WIN IgE study had no specific funding (it is an ancillary study of the WIN Lupus trial, which was funded by the French Ministry of Health (PHRC 2010)).

  • Competing interests MH, LC and KB declare they have no competing interests. NC declares consulting fees from Argenx. NJ-C declares consulting fees from GSK and Otsuka; lecture fees from GSK and Otsuka; payment for expert testimony from Otsuka; support for attending meetings and travel from Sanofi. ED declares consulting fees from GSK, Astra Zeneca, Amgen, Otsuka and Novartis; lecture fees from GSK and Astra Zeneca; support for attending meetings and travel from GSK, Otsuka, Astra Zeneca Alexion; participation on a DSMB for Alexion.

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

  • Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.