Article Text

Lower disease activity but higher risk of severe COVID-19 and herpes zoster in patients with systemic lupus erythematosus with pre-existing autoantibodies neutralising IFN-α
  1. Alexis Mathian1,2,
  2. Paul Breillat2,
  3. Karim Dorgham2,
  4. Paul Bastard3,4,5,6,
  5. Caroline Charre7,8,
  6. Raphael Lhote1,
  7. Paul Quentric2,
  8. Quentin Moyon9,
  9. Alice-Andrée Mariaggi7,8,
  10. Suzanne Mouries-Martin10,
  11. Clara Mellot2,
  12. François Anna11,
  13. Julien Haroche9,
  14. Fleur Cohen-Aubart9,
  15. Delphine Sterlin2,12,
  16. Noël Zahr13,
  17. Adrian Gervais3,4,
  18. Tom Le Voyer3,4,
  19. Lucy Bizien3,
  20. Quentin Amiot12,
  21. Micheline Pha1,
  22. Miguel Hié1,
  23. Francois Chasset14,
  24. Hans Yssel2,
  25. Makoto Miyara2,12,
  26. Pierre Charneau11,
  27. Pascale Ghillani-Dalbin12,
  28. Jean-Laurent Casanova3,4,5,6,15,
  29. Flore Rozenberg7,
  30. Zahir Amoura2,9,
  31. Guy Gorochov2,12
  1. 1Assistance Publique–Hôpitaux de Paris (AP-HP), Groupement Hospitalier Pitié–Salpêtrière, Centre de Référence pour le Lupus, le Syndrome des anti-phospholipides et autres maladies auto-immunes rares, Service de Médecine Interne 2, Institut E3M, Paris, France
  2. 2Sorbonne Université, Inserm, Centre d’Immunologie et des Maladies Infectieuses (CIMI-Paris), Paris, France
  3. 3Laboratory of Human Genetics of Infectious Diseases, Necker Branch, Inserm U1163, Necker Hospital for Sick Children, Paris, France
  4. 4University of Paris Cité, Imagine Institute, Paris, France
  5. 5St. Giles Laboratory of Human Genetics of Infectious Diseases, Rockefeller Branch, The Rockefeller University, New York, NY, USA
  6. 6Department of Pediatrics, Necker Hospital for Sick Children, Paris, France
  7. 7Université de Paris Cité, Assistance Publique–Hôpitaux de Paris, Hôpital Cochin, Service de Virologie, Paris, France
  8. 8INSERM U1016, CNRS UMR8104, Institut Cochin, Paris, France
  9. 9Sorbonne Université, Assistance Publique–Hôpitaux de Paris (AP-HP), Groupement Hospitalier Pitié–Salpêtrière, Centre de Référence pour le Lupus, le Syndrome des anti-phospholipides et autres maladies auto-immunes rares, Service de Médecine Interne 2, Paris, France
  10. 10Centre Hospitalier Universitaire de Dijon, Hôpital François-Mitterrand, service de médecine interne et maladies systémiques (médecine interne 2), Dijon, France
  11. 11Pasteur-TheraVectys Joint Lab, Institut Pasteur, Paris, France
  12. 12Département d'Immunologie, AP-HP, Groupement Hospitalier Pitié–Salpêtrière, Paris, France
  13. 13Service de Pharmacologie, Assistance Publique-Hôpitaux de Paris, Groupement Hospitalier Pitié-Salpêtrière, Paris, France
  14. 14Sorbonne Université, Service de dermatologie et allergologie, hôpital Tenon, AP-HP, Paris, France
  15. 15Howard Hughes Medical Institute, New York, NY, USA
  1. Correspondence to Professor Guy Gorochov; guy.gorochov{at}


Objectives Type-I interferons (IFNs-I) have potent antiviral effects. IFNs-I are also overproduced in patients with systemic lupus erythematosus (SLE). Autoantibodies (AAbs) neutralising IFN-α, IFN-β and/or IFN-ω subtypes are strong determinants of hypoxemic COVID-19 pneumonia, but their impact on inflammation remains unknown.

Methods We retrospectively analysed a monocentric longitudinal cohort of 609 patients with SLE. Serum AAbs against IFN-α were quantified by ELISA and functionally assessed by abolishment of Madin-Darby bovine kidney cell protection by IFN-α2 against vesicular stomatitis virus challenge. Serum-neutralising activity against IFN-α2, IFN-β and IFN-ω was also determined with a reporter luciferase activity assay. SARS-CoV-2 antibody responses were measured against wild-type spike antigen, while serum-neutralising activity was assessed against the SARS-CoV-2 historical strain and variants of concerns.

Results Neutralising and non-neutralising anti-IFN-α antibodies are present at a frequency of 3.3% and 8.4%, respectively, in individuals with SLE. AAbs neutralising IFN-α, unlike non-neutralising AAbs, are associated with reduced IFN-α serum levels and a reduced likelihood to develop active disease. However, they predispose patients to an increased risk of herpes zoster and severe COVID-19 pneumonia. Severe COVID-19 pneumonia in patients with SLE is mostly associated with combined neutralisation of different IFNs-I. Finally, anti-IFN-α AAbs do not interfere with COVID-19 vaccine humoral immunogenicity.

Conclusion The production of non-neutralising and neutralising anti-IFN-I antibodies in SLE is likely to be a consequence of SLE-associated high IFN-I serum levels, with a beneficial effect on disease activity, yet a greater viral risk. This finding reinforces the recommendations for vaccination against SARS-CoV-2 in SLE.

  • COVID-19
  • lupus erythematosus, systemic
  • cytokines
  • inflammation
  • autoimmunity

Data availability statement

Data are available upon reasonable request.

This article is made freely available for personal use in accordance with BMJ’s website terms and conditions for the duration of the covid-19 pandemic or until otherwise determined by BMJ. You may use, download and print the article for any lawful, non-commercial purpose (including text and data mining) provided that all copyright notices and trade marks are retained.

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.


  • Anti-interferon (IFN)-α autoantibodies (AAbs) have been reported in 5%–27% of patients with systemic lupus erythematosus (SLE), it is, however, as yet unclear whether their occurrence is pathogenic, protective or a reflection of a general tendency towards autoreactivity.


  • Neutralising and non-neutralising anti-IFN-α AAbs are present at a frequency of 3.3% and 8.4%, respectively, in patients with SLE.

  • AAbs neutralising IFN-α are associated with reduced IFN-α serum levels and a reduced likelihood to develop active disease.

  • AAbs neutralising IFN-α are associated with a history of severe COVID-19 pneumonia and episodes of cutaneous herpes zoster.


  • Monitoring anti-IFN-α antibodies in patients with SLE can help identify patients at risk of developing serious viral infections.


Type-I interferons (IFNs-I) play a central role in the early control of viral infections. Inborn errors of IFN-I immunity were recently found in patients with life-threatening COVID-19.1 2 Autoantibodies (AAbs) neutralising IFNs-I were also found in 7% and 15% of patients with severe and critical COVID-19 pneumonia, respectively.3–6 They were also found in about a third of a cohort of patients with yellow fever vaccine-associated disease.7 However, little is known about the circumstances in which neutralising AAbs directed at IFNs-I appear and whether they might also have anti-inflammatory effects. The IFN family of cytokines is indeed involved in systemic lupus erythematosus (SLE) pathogenesis, an autoimmune disease affecting mostly young women and where persistent overexpression of IFNs-I, notably IFN-α, is observed.8 While anti-IFN-α AAbs have been reported in 5% to 27% of patients with SLE,9–12 it is, however, as yet unclear whether the occurrence of these AAbs in the context of SLE is pathogenic, protective or a reflection of a general tendency towards autoreactivity. It has been suggested that endogenous anti-IFN-α AAbs may have a regulatory, protective, role against disease activity.10 11 However, it is difficult to draw firm conclusions from these studies involving only small numbers of patients. Indeed, if the presence of anti-IFN-α AAbs has reportedly been associated with reduced downstream IFN pathway activity in patients with SLE, it was either not11 or only weakly10 associated with a decrease in disease activity. Anti-IFN-α antibodies were previously described in two patients with SLE with severe COVID-19,13 but their clinical impact on SLE activity was not explored. Furthermore, although targeting IFN-I signalling pathways represents a promising therapeutic approach for SLE, as evidenced by the recent approval of the IFN-I receptor antagonist anifrolumab by the US Food and Drug Administration14 and the European Medicines Agency,15 the potential long-term viral risk caused by this type of treatment is of concern.

In the present study, we retrospectively analysed immunological and clinical data in a monocentric longitudinal cohort of 609 patients with SLE and focused on the association between the presence and the neutralisation capacity of serum anti-IFN-α AAbs, infectious complications and disease evolution. We hypothesised that neutralising anti-IFN-α AAbs might confer an additional viral risk to patients with SLE but could also have a disease-ameliorating effect.

Patients, materials and methods

Study design and patients

The retrospective longitudinal study reported here was conducted between June 2006 and June 2021 at the French National Referral Center for SLE and Antiphospholipid Antibody Syndrome and Other Autoimmune Disorders, Paris, France, regrouping out or inpatients with active or quiescent, untreated or treated disease. Serum samples were randomly obtained from patients diagnosed with SLE according to the 1997 American College of Rheumatology criteria for SLE classification or the 2019 European League Against Rheumatism/American College of Rheumatology classification criteria for SLE.16 17 Patients seen in outpatient clinic or during hospital care were randomly included in the study, regardless of disease activity and treatment. Serum samples were kept frozen until anti-interferon-α AAbs were assessed. See online supplemental file for the designs of the clinical studies. The study was approved by the ethical committee of Sorbonne Université (CER2020-012, CER2021-011 and CER2021-099) and informed consent was obtained from all participants.

Supplemental material

Measurement of anti-IFN-α AAbs

Auto-Abs against IFN-α were quantified using the anti-IFN-α Antibody Human ELISA Kit (Thermo Fisher, Invitrogen), according to the manufacturer’s instructions. The positivity threshold of the assay was 15 ng/mL.

Determination of biological activity of IFN-α by IFN-α bioassay

Serum IFN-α biological activity was determined by assessing the protection conferred by each patient’s serum to cultured Madin-Darby bovine kidney (MDBK) cells challenged with vesicular stomatitis virus (VSV), as previously described.18–21 Serum IFN-α levels are expressed in IU/mL after comparison with IFN-α2b reference (Introna, Shering Plough), standardised against the National Institutes of Health reference Ga 023-902-530 titrated under the same conditions as the SLE patients’ serum samples. The lower limit of detection was 2 IU/mL. Serum IFN-α activity in healthy individuals is undetectable (ie,<2 IU/mL).22 23

Functional evaluation of anti-IFN-α AAbs by VSV assay

The blocking activity of anti-IFN-α AAb-containing serum was assessed as previously described.24 Neutralisation experiments were performed by the titration of serial dilutions of serum positive for anti-IFN-α AAbs against 10 IU/mL (50 pg/mL) of IFN-α2b (Introna, Shering Plough), following the previously described antiviral assay. Serum and IFN-α were incubated together for 30 min at room temperature before being added to MDBK cells. End points were scored at 50% cytopathic effect (CPE). Sera to be tested for their anti-IFN-α neutralisation capacity were previously inactivated at 56°C for 60 min to remove endogenous IFN-α activity. Neutralising titres correspond to the serum dilution at 50% CPE × 10. For clinical studies, only sera with neutralisation titres >30 were considered significant.

Functional evaluation of anti-IFN-I AAbs by luciferase reporter assay

The blocking activity against IFN-α2 and IFN-ω at 102 pg/mL and 104 pg/mL, and IFN-β at 104 pg/mL were determined with a reporter luciferase activity assay as previously described.4

SARS-CoV-2 serological analysis

Serum levels of SARS-CoV-2-specific immunoglobulin G (IgG) antibodies were assessed using an ELISA specific for antinucleocapsid IgG (Euroimmun, France) or the Maverick SARS-CoV-2 Multi-Antigen Serology Panel (Genalyte, USA), according to the manufacturer’s instructions, as previously described.25 The latter is designed to detect antibodies specific for five SARS-CoV-2 antigens: nucleocapsid, spike S1 receptor-binding domain (RBD), spike S1S2, spike S2 and spike S1, within a multiplex format based on photonic ring resonance technology.

SARS-CoV-2 pseudoneutralisation assay

Lentiviral particles carrying the luciferase Firefly gene and pseudotyped with spikes of SARS-CoV-2 historical strain or variants of concerns (VOCs were produced by triple transfection of 293 T cells as previously described.25

Statistical analysis

Qualitative variables are expressed as number (%) and quantitative variables as the mean±SD or median (quartiles), as appropriate. The Mann-Whitney U-test or Student’s t test for continuous data and Fisher’s exact or χ2 test for categorical data were used to compare independent groups. Spearman’s correlation coefficients were computed for quantitative values. The diagnostic performance of the serum anti-IFN-α AAb levels as assessed by ELISA, to detect an IFN-α-neutralising capacity, was investigated by analysing receiver operating characteristic (ROC) curves, with the capacity to neutralise 10 IU/mL of IFN-α serving as the gold standard. The areas under the ROC curves (AUCs) to differentiate sera with IFN-α-neutralising capacity versus sera without were calculated. The optimal threshold was determined using a compromise among the minimum sensitivity–specificity difference and the Youden’s index. We measured the statistical association between the occurrence of severe or critical COVID-19 pneumonia in patients with SLE and different sets of neutralising anti-IFN-I capacities. Time to flare was studied by the mean of Kaplan-Meier method and compared using Log-Rank tests for patients in whom immunosuppressive and corticosteroid therapy were not increased on the day monitoring was initiated. We performed a sensitivity analysis also including patients in whom immunosuppressive or corticoid therapy was increased on the day monitoring was initiated. Crude HRs were calculated using the Log-Rank or Mantel-Haenszel estimate when appropriate. All tests were two sided and p<0.05 defined significance. Statistical analyses were performed using GraphPad Prism, V.8.0.1 software (GraphPad Software, San Diego, California), R software, V.3.6.3 and V.4.0.5 and the web tool easy ROC, V.


High prevalence of neutralising and non-neutralising anti-IFN-α AAbs in SLE

The presence of serum anti-IFN-α AAbs was detected by ELISA in 71 (11.7%) of the 609 patients we analysed, with levels measured at least once above 500 ng/mL in 27 (38.0%) patients and were usually persistent, since they became undetectable in only 10 out of 63 (16%) patients followed for a median (IQR) time of 4.2 years (3.6–6.4) (online supplemental figure 1). There was no significant difference in terms of gender or median age between patients with ELISA-detectable anti-IFN-α AAbs (aIFN-α+) or not (aIFN-α): 65 out of 71 aIFN-α+ patients (91.5%) versus 509 out of 538 aIFN-α patients (94.6%) were women, p=0.28 and 34.6 (26.5–46.5) years versus 37.7 (29.5–49.4), p=0.06, respectively). We then assessed the biological activity of these AAbs. Only 20 (28.2%) of the 71 sera with ELISA-detectable anti-IFN-α AAbs significantly abolished MDBK cell protection by IFN-α2 against viral challenge. Neutralisation capacity was proportional to anti-IFN-α AAb levels (figure 1A,B), although some rare serum samples containing high AAb levels were not endowed with neutralising activity (figure 1A). The AUC for anti-IFN-α AAb serum levels, differentiating between IFN-α-neutralising and non-neutralising sera, was 0.90 (95% CI 0.85 to 0.96, figure 1C), the optimal ELISA threshold for prediction of neutralisation activity, as determined using the minimum sensitivity–specificity difference and the Youden’s index, being 310 ng/mL. Proportions of patients with neutralising activity were similar in all age groups (figure 2A). In conclusion, not all anti-IFN-α AAbs have neutralisation potential. Although evaluation of serum-neutralising activity remains the gold standard, simple assessments with ELISA assays are informative since a strong correlation with biological activity was observed.

Figure 1

Neutralising and non-neutralising anti-IFN-α AAbs in SLE. (A) IFN-α neutralisation potential contained in 126 serum samples from 71 SLE patients with anti-IFN-α AAbs, measured using the MDBK antiviral activity cell assay. Each vertical bar represents a serum sample. Samples are distributed along the x-axis according to the increasing serum level of anti-IFN-α AAbs. Optimal cut-off point of anti-IFN-α AAb serum concentration, associated with IFN-α neutralising capacity (310 ng/mL), as determined using the minimum sensitivity—specificity difference and the Youden’s index is indicated (horizontal dashed grey line). (B) Correlation between anti-IFN-α AAb serum concentrations and serum neutralisation titres. Each dot represents an individual. Only neutralising samples were analysed (n=60). Spearman’s rank correlation coefficient was used. (C) Diagnostic performance of serum anti-IFN-α AAbs measured by ELISA to predict neutralisation of 10 IU/mL (50 pg/mL) of IFN-α2. Area under receiver operating characteristics (ROC) curve (AUC) is indicated. The optimal cut-off point (red arrow), determined using the minimum sensitivity–specificity difference and the Youden’s index is represented. IFN, interferon; MDBK, Madin-Darby bovine kidney; SLE, systemic lupus erythematosus.

Figure 2

Anti-IFN-α AAbs and viral infections in SLE. (A) Serum anti-IFN-α AAb levels, as determined by ELISA, in SLE patients (n=609) according to age. Indicated proportions of IFN-α neutralisation activity were assessed using the MDBK cell assay. (B) History of viral infections in relation with neutralisation activity of serum anti-IFN-α AAbs. P values were calculated using the Fisher’s exact test. p<0.05 was considered significant. *p<0.05 and ***p<0.001. CIL/CIN/CC, cervical intraepithelial lesions or cervical intraepithelial neoplasia or cervical cancer; IFN, interferon; SLE, systemic lupus erythematosus.

Anti-IFN-α-neutralising AAbs are associated with increased viral risk in SLE

We next searched for comorbidities associated with the presence of anti-IFN-α AAbs in SLE. In order to analyse the impact of anti-IFN-α AAbs on the risk of viral infection in SLE, we designed a retrospective cohort study in which all patients with SLE with anti-IFN-α AAbs (aIFN-α+) were compared with patients without anti-IFN-α AAbs (aIFN-α) at a 1:2 ratio (see online supplemental patients, materials and methods). While none of the aIFN-α patients experienced a severe COVID-19 pneumonia, five patients (7%) out of the 71 aIFN-α+ patients developed severe or critical COVID-19 pneumonia (table 1). The presence of anti-IFN-α-neutralising AAbs, unlike that of non-neutralising AAbs, was associated in a statistically significant manner with a history of severe or critical COVID-19 pneumonia, episodes of cutaneous herpes zoster and severe viral infection (p=3.10−4, p=0.03 and p=10−4, respectively, figure 2B and online supplemental table 2). Of note, the eight cases of severe viral infections in patients with anti-IFN-α-neutralising AAbs included five cases of COVID-19 pneumonia, two cutaneous herpes zoster and one varicella pneumonia. Importantly, patients had samples collected before SARS-CoV-2 infection, and anti-IFN-α AAbs were detected in all cases, prior to infection, further suggesting that they are a cause, rather than a consequence, of severe viral infection. On the other hand, aIFN-α+ patients were not at higher risk to suffer from warts and human papillomavirus (HPV)-induced cervical lesions, as suggested by previous genetic studies on predisposition to HPV infection.27

Table 1

Demographics, IFN-I neutralising capacities and severity of SARS-CoV-2 infection in 17 patients with SLE tested positive for circulating serum anti-IFN-α AAbs

Combined neutralisation of different IFN-I subtypes is associated with severe COVID-19

Given that in the general population, as well as in SLE patients, anti-IFN-α AAbs are frequently associated with the presence of antibodies against other IFNs-I, such as IFN-β and IFN-ω,3 4 9 11 we tested whether their coexistence was associated with an increased infectious risk. Serum sampled as close as possible to the COVID-19 pandemic onset were assessed for their neutralisation capacity against IFN-α, and IFN-ω at 102 pg/mL and IFN-β at 104 pg/mL using a luciferase assay, as previously described.4 None of the 134 sera lacking detectable levels of anti-IFN-α AAbs was able to neutralise IFN-α2 or IFN-β, and only 4 (3%) neutralised IFN-ω. In contrast, neutralising activities against IFN-α2, IFN-β and IFN-ω were more frequently detected (18 (25%), 12 (17%) and 15 (21%) sera, respectively) in the 71 sera with ELISA-detectable anti-IFN-α AAbs. A total of 30 (42%) of the 71 aIFN-α+ sera tested neutralised at least one IFN-I, while 9 (13%) and 3 (4%) neutralised two and three IFNs-I, respectively. A high concentration of anti-IFN-α AAbs was associated with an increasing number of IFN-I neutralising abilities. Indeed, anti-IFN-α AAb concentrations in serum which neutralised at least two IFNs-I (median (Q1–Q3); 5592 (837–70 175) ng/mL) were significantly higher than those in serum which neutralised a single IFN-I (350 (72–2485) ng/mL; p=0.009) and in serum which did not neutralise IFN-I (53 (32–154) ng/mL; p<10−4). Anti-IFN-α AAb concentrations in serum of these latter groups also differed significantly (p=0.008).

Importantly, the occurrence of severe or critical COVID-19 was significantly associated with the neutralisation of IFN-α2 or IFN-ω (p=0.013 and p=0.005, respectively, table 2). Finally, the analysis confirmed that severe or critical COVID-19 in SLE was very significantly associated with combined neutralisation of both IFN-α2 and IFN-ω subtypes (p<10−4, table 2), as recently observed in the general population.28 Of note, the only patients with SLE in this cohort who deceased of COVID-19 had AAbs that neutralised all three IFN-I subtypes tested, suggesting that the severity of COVID-19 pneumonia is even higher in individuals neutralising several IFN-I.28 It should also be noted that two of the five patients who experienced a severe COVID-19 presented comorbidities conditions such as obesity, immunosuppressive therapy and renal allograft (table 1).

Table 2

Risk of severe or critical COVID-19 pneumonia in patients with SLE, carrying different sets of neutralising IFN-I activities

Anti-IFN-α-neutralising AAbs are associated with reduced SLE disease activity

We compared the clinical course of SLE in the presence or absence of anti-IFN-α AAbs (see online supplemental patients, materials and methods). Patients with neutralising anti-IFN-α AAbs had reduced disease activity, less flares and less clinically active SLE, were more likely to be in remission or in lupus low disease activity states compared with patients who lacked neutralising anti-IFN-α AAbs (figure 3A). Biological markers of SLE disease activity, such as elevated antidouble-stranded DNA Ab serum levels (ie, Farr assay), decrease in complement component C3 and increase in serum IFN-α levels were also reduced in patients with neutralising anti-IFN-α AAbs compared with patients without (figure 3A). Other characteristics of lupus disease were similar between the two groups (online supplemental table 3). Non-neutralising anti-IFN-α AAbs were associated with higher IFN-α serum levels and the presence of anti-RNP and anti-Sm Abs. Of the 18 patients with neutralising anti-IFN-α AAbs in whom immunosuppressive and corticosteroid therapy were not increased, none experienced a lupus flare during the following year (figure 3B). Log-Rank test analysis showed a significantly higher risk of relapse in patients with non-neutralising anti-IFN-α AAbs, as compared with patients with neutralising anti-IFN-α AAbs (HR 4.78 (95% CI 1.02 to 22.40), p=0.047). The results from a sensitivity analysis, including patients in whom immunosuppressive or corticoid therapy was increased at the beginning of the follow-up, showed that only one patient out of 20 with neutralising anti-IFN-α AAbs experienced a lupus flare during the following year. In summary, non-neutralising anti-IFN-α AAbs are more prevalent and are typically associated with both unstable disease and high IFN-α serum levels. In contrast, the presence of neutralising AAbs in patients with SLE was associated with a concomitant reduction in levels of serum IFN-α and disease activity.

Figure 3

SLE disease activity and BNT162b2 vaccine immunogenicity. (A) SLE activity assessed with the SLEDAI-2K score (left), clinical and biological markers of SLE disease activity (middle) and IFN-α serum levels (right) according to anti-IFN-α AAb status. Left and right, columns represent the mean values of disease activity and IFN-α serum levels and vertical lines show positive SD. (B) Kaplan-Meyer analysis of the risk to develop SLE flares in relation to baseline anti-IFN-α AAb status. Red, neutralising aIFN-α+; blue, non-neutralising aIFN-α+ (positivity ELISA threshold: 15 ng/mL); grey, aIFN-α−. Vertical ticks indicate patients who remained flare-free but did not have a full year of clinical follow-up (censored data). Curves were compared using Log-Rank tests. Crude hazard ratios (HR) were calculated. P<0.05 was considered significant. (C) BNT162b2-vaccinated patients (two injections) evaluated at day 42 after first injection. Left, comparison of anti-RBD IgG serum levels measured by photonic ring immunoassay in patients with (n=9) and without (n=19) serum anti-IFN-α AAbs. Pink solid circles and empty circles represent IFN-I-neutralising and IFN-I-non-neutralising aIFN-α+ patients, respectively. Median values, first and third quartiles, are indicated. P values were calculated using the Mann-Whitney U test. Right, serum with (n=10) or without (n=19) anti-IFN-α AAbs tested for neutralisation of D614G SARS-CoV-2 and variants B.1.1.7 (alpha), B.1.351 (beta), B.1.1.28 (gamma) and B.1.617.2 (delta). Patients were defined as ‘non-neutralisers’ or ‘neutralisers’ according to the absence or presence of neutralising activity at first serum dilution (1/30). The Mann-Whitney U test for continuous variables and the Fisher’s exact test for categorical variables were used for bivariable analysis. p<0.05 was considered significant. *p<0.05. IFN-α AAbs, anti-interferon-alpha autoantibodies; SLE, systemic lupus erythematosus.

Anti-IFN-α AAbs do not interfere with COVID-19 vaccine efficacy

Vaccination currently represents the best option to prevent serious infections in patients with SLE. We reasoned that neutralisation of IFN-α signalling might possibly dysregulate IFN-dependent B cell responses29 and limit vaccine-induced antibody production. In order to determine whether anti-IFN-α AAbs could interfere with COVID-19 vaccine efficacy, we performed a subanalysis of the results we recently obtained in a cohort of patients with SLE,30 evaluating their SARS-CoV-2-specific immune responses after BNT162b2 vaccination in presence or absence of these AAbs. IFN-I-neutralising activity was confirmed in 50% of the 10 vaccinated aIFN-α+ patients tested, whereas demographics and main bioclinical characteristics were similar in aIFN-α+ and aIFN-α patients (online supplemental table 4). Vaccine-induced anti-SARS-CoV-2 spike RBD IgG levels, and serum-neutralising capacity of SARS-CoV-2 and its major variants, were similar in both groups, thus confirming that aIFN-α+ patients are able to mount an efficacious anti-SARS-CoV-2 humoral vaccine response, similar to that of aIFN-α patients (figure 3C). In conclusion, although only a limited number of vaccinated patients with SLE could be analysed, the results nevertheless show that anti-IFN-α AAbs do not seem to interfere with COVID-19 humoral vaccine response.


The COVID-19 outbreak has illustrated the fact that a previously poorly recognised form of autoimmunity underlies some severe forms of COVID-19 disease,3–7 although the mechanisms driving the appearance of the anti-IFN-I AAbs and their potential broader medical impact remain unknown. Besides reported SLE-associated cases,9–12 31 these AAbs have also been found in patients with thymoma,32 myasthenia gravis33 34 or affected by various primary immune deficiencies.35–39 However, their potential inflammatory disease-ameliorating effects until now remained unexplored.

Here, we analysed a longitudinal cohort of 609 patients with SLE, a disease driven by IFN-α, evolving by successive phases of relapses and remissions affecting from 29 to 367 per 100 000 individuals in North America and Europe.40 We show that the prevalence of anti-IFN-α antibodies is particularly elevated in this population. As expected, we confirm that this novel form of autoimmunity is associated with a greater risk to contract severe COVID-19 disease. We also highlight its association with herpes zoster. It should be emphasised that AAbs directed to human IFN-α were first observed in a patient with varicella-zoster disease,41 but that link had been not confirmed as yet. More recently, the administration of anifrolumab, a human monoclonal antibody that binds IFN-I receptor subunit, was associated with an increased incidence of herpes zoster,42 which confirms that IFN-I blockade impairs varicella-zoster recurrences control. Unlike others,43 we did not observe reactivation of either type 1 and 2 herpes simplex virus or cytomegalovirus in patients with anti-IFN-I AAbs. We also show that IFN-α autoimmunity appears to have a beneficial effect on inflammatory disease activity.

The analysis of this cohort of patients with SLE might provide some clues regarding the mechanism underlying the development of anti-IFN-I AAbs. Overall, the results suggest that abnormally elevated IFN-I levels elicit an AAb response that eventually matures from non-neutralising to neutralising in some patients with SLE. This evolution might be predicted from our observation of two distinct clinical presentations associated with anti-IFN-I AAbs; either, (1) elevated IFN-I levels, instable SLE disease and non-neutralising anti-IFN-I AAbs or (2) low IFN-I levels, quiescent SLE disease and neutralising anti-IFN-I AAbs. This interpretation is in line with the observation that patients treated with IFN-α or IFN-β are also prone to develop AAbs targeting these cytokines.44–46 Future longitudinal studies will be necessary to explore the relationship between neutralisation activity and somatic hypermutation-driven molecular evolution that may underlie in vivo promotion of neutralising anti-IFN-I AAbs.

Our study also has immediate implications in terms of medical management: (1) considering their prevalence in SLE, affected patients should be screened for the presence of anti-IFN-I AAbs, (2) because the biological activity of these AAbs, is correlated with their serum concentration, their mere titration might, in most instances, inform on their clinical relevance, (3) since anti-COVID-19 vaccination is well tolerated in SLE30 and since its efficacy is not impaired by anti-IFN-I AAbs, patients with SLE carrying these AAbs should be vaccinated against COVID-19 as a priority and (4) preventive and/or early curative antiviral treatment47 should also be considered in cases of SARS-CoV-2 infection in patients with SLE with serum anti-IFN-I AAbs. Finally, our results have also implications regarding innovative therapeutic options that are currently being tested in SLE.48 Because viral risk seems likely associated with the neutralisation of more than one IFN-I subtype, we would argue that anti-IFN intervention in SLE and other diseases might not concomitantly target all IFNs. Long-term placebo-controlled assessment of patients treated with anifrolumab, that interferes with all IFNs-I besides IFN-α, was recently reported.49 A total of seven deaths were attributed to infections (four pneumonia and three COVID-19) in anifrolumab-treated subjects, as compared with none in the group of patients receiving placebo.49 The interpretation of these data should, however, take into account the large number of patients treated with anifrolumab, compared with those receiving placebo as well as the fact that the observation period spanned the first year of the pandemic prior to vaccination and implementation of effective treatments for severe COVID-19. Our own study also dates back to the prevaccination era of the pandemic and none of the patients who developed severe or critical COVID-19 in our cohort had been vaccinated against SARS-CoV-2. The forthcoming anifrolumab safety data collected in patients vaccinated against SARS-CoV-2 should provide more important insights.

The main limitation of our study is associated with its design that was limited to a retrospective analysis of clinical data. However, there is arguably no reason to expect that clinical flares would tend to be better recorded in one group of patients or the other, characterised by the presence or absence of anti-IFN-α AAbs, because this biomarker was never recorded prior to the present study, and, therefore, had no impact on medical care. An additional limitation, pertaining to the estimation of viral risk, was study size. Even in a study that comprised several hundred patients affected by a rare disease, cases that present both anti-IFN-α AAbs and a history of COVID-19 constitute only a small subset. As a result, only few severe or critical COVID-19 cases were recorded, but it was nevertheless possible to establish a significative link between presence of AAbs against IFN-I and COVID-19 severity, furthermore taking into account that the majority of patients with SLE are women, often young, and, therefore, at lower risk of severe infection. It should also be underlined that the link between anti-IFNs-I and COVID-19 has been confirmed in different studies, including a cohort of 3595 patients hospitalised with critical COVID-19 pneumonia.4 5 50–59 Our study setup was not designed to estimate the prevalence of anti-IFN-α AAbs among patients with SLE with severe COVID-19 pneumonia. Other factors will obviously contribute to an enhanced risk of developing a severe COVID-19, as suggested by the presence of associated comorbidities in two out of the five patients with anti-IFN-α AAbs who developed a severe COVID-19 in the cohort.60 Finally, although we report that the presence of neutralising anti-IFN-α AAbs did not interfere with the induction of vaccine-induced antibody responses, we could not analyse the effect of these AAbs on the development of SARS-CoV-2-specific T cell immunity, and this point will, therefore, require further study since it was recently reported that a small proportion of individuals with such AAbs might not be fully protected by the vaccine.61 A final limitation, which is not addressed here, is associated with the genetic evolution of SARS-CoV-2, which may alter its IFN-I sensitivity.

In summary, while neutralising anti-IFN-I AAbs seem to confer increased viral susceptibility, they are also associated with reduced SLE disease activity. It is tempting to not only speculate that immunisation against IFN-α could be a consequence of elevated levels of this cytokine recurrently observed in patients with SLE with active disease, but also that neutralising anti-IFN-I autoimmunity is progressively acquired in these patients.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Ethics approval

This study involves human participants and was approved by ethical committee of Sorbonne Université (CER2020-012, CER2021-011 and CER2021-099). Participants gave informed consent to participate in the study before taking part.


We thank the patients, the nurses and the Department of Internal Medicine 2 staff who participated in this study, Laura Wakselman, Naima Zemirli and Juliette Blondy from clinical research unit (URC) of Pitié–Salpêtrière hospital for helping with regulatory and ethical issues. We warmly thank the members of both branches of the Laboratory of Human Genetics of Infectious Diseases for discussions and Y. Nemirovskaya, M. Woollet, D. Liu, S. Boucherit, C. Rivalain, M. Chrabieh and L. Lorenzo for administrative assistance.


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.


  • AM, PB and KD are joint first authors.

  • ZA and GG are joint last authors.

  • Handling editor Josef S Smolen

  • AM, PB and KD contributed equally.

  • ZA and GG contributed equally.

  • Contributors AM, PBr, KD, PBa, SM-M, CM, RL, JLC, FR, ZA and GG contributed to the conception and design of the study; AM, PBr, KD, PBa, CC, RL, QM, AAM, SMM, CM, FA, JH, FCA, DS, NZ, AG, TLV, LB, QA, MP, MH, FC, MM, PGD, FR and ZA were involved in the acquisition of data; AM, PBr, KD, PBa, CC, RL, PQ, QM, AAM, SMM, CM, FA, DS, HY, PC, PGD, JLC, FR, ZA and GG contributed to the analysis and interpretation of data. All authors contributed to drafting and/or revising the manuscript. AM acts as guarantor for the overall content of the study.

  • Funding The study was supported by the Recherche Hospitalo-Universitaire RHU-COVIFERON project under the program 'Investissement d’Avenir' launched by the French Government and implemented by the Agence Nationale de la Recherche (ANR) with the reference ANR-21-RHUS-08 and by the EU Horizon 101057100 UNDINE project (JLC and GG). The Laboratory of Human Genetics of Infectious Diseases is supported by the Howard Hughes Medical Institute, the Rockefeller University, the St. Giles Foundation, the National Institutes of Health (NIH) (R01AI088364 and R01AI163029), the National Center for Advancing Translational Sciences (NCATS), the NIH Clinical and Translational Science Award (CTSA) program (UL1 TR001866), a Fast Grant from Emergent Ventures, Mercatus Center at George Mason University, the Yale Center for Mendelian Genomics and the GSP Coordinating Center funded by the National Human Genome Research Institute (NHGRI) (UM1HG006504 and U24HG008956), the Yale High-Performance Computing Center (S10OD018521), the Fisher Center for Alzheimer’s Research Foundation, the Meyer Foundation, the French National Research Agency (ANR) under the 'Investments for the Future' program (ANR-10-IAHU-01), the Integrative Biology of Emerging Infectious Diseases Laboratory of Excellence (ANR-10-LABX-62-IBEID), the French Foundation for Medical Research (FRM) (EQU201903007798), the FRM and ANR GENCOVID project, the ANRS-COV05, ANR GENVIR (ANR-20-CE93-003) and ANR AABIFNCOV (ANR-20-CO11-0001) projects, the European Union’s Horizon 2020 research and innovation program under grant agreement on. 824110 (EASI-genomics), the Square Foundation, Grandir—Fonds de solidarité pour l’enfance, the Fondation du Souffle, the SCOR Corporate Foundation for Science, Institut National de la Santé et de la Recherche Médicale (INSERM), The French Ministry of Higher Education, Research, and Innovation (MESRI-COVID-19) and the University of Paris. PBa was supported by the French Foundation for Medical Research (FRM, EA20170638020) and by the MD-PhD program of the Imagine Institute (with the support of Fondation Bettencourt-Schueller). PBr was supported by the Regional Health care Agency of Île-de-France (bourse année recherche de l’Agence Régional de Sante) and the Villa M grant (with the support of Groupe Pasteur Mutualité Hospitalier).

  • Competing interests AM has received grant/research support from Sobi; participated in advisory board related to lupus for AstraZeneca; received payment for expert testimony for GSK; received support for attending meetings and/or travel from AstraZeneca and GSK; received consulting fees, speaking fees and honoraria from AstraZeneca and GSK. FC has received grant/research support from AstraZeneca; participated in advisory board related to lupus for AstraZeneca, GSK, Celgene and Principabio; received speaking fees and honoraria from AstraZeneca and GSK. ZA has received grant/research support from GSK, AstraZeneca, Roche, Novartis, Amgen; participated in advisory board related to lupus for GSK, AstraZeneca, Kezar, Amgen, Otsuka; received consulting fees, speaking fees and honoraria from AstraZeneca and GSK.

  • Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.

  • 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.