IL-1RA Antibodies in Myocarditis after SARS-CoV-2 Vaccination.

To the Editor: Myocarditis associated with messenger RNA (mRNA) vaccines against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) predominantly affects male adolescents and young male adults (14 to <30 years of age) and typically occurs after receipt of the second vaccine dose.[1,2] In adults with critical coronavirus disease 2019 (Covid-19) and in cases of multisystem inflammatory syndrome in children (MIS-C), we recently discovered neutralizing autoantibodies targeting the endogenous interleukin-1 receptor antagonist (IL-1RA), which inhibits interleukin-1 signaling and inflammation.[3,4]

In this study, we evaluated the prevalence of antibodies neutralizing IL-1RA and progranulin, which inhibits tumor necrosis factor signaling, in 69 patients (14 to 79 years of age) who had clinically suspected myocarditis after SARS-CoV-2 vaccination. A total of 61 patients underwent endomyocardial biopsy.

Myocarditis was confirmed by biopsy in 40 of 61 patients (Figure 1A). Among patients with histologically confirmed myocarditis, anti–IL-1RA antibodies were found in 9 of 12 patients (75%) younger than 21 years of age, as compared with 3 of 28 patients (11%) 21 years of age or older. Anti–IL-1RA antibodies were not detectable in the 21 patients in whom biopsy ruled out the diagnosis of myocarditis (Figure 1B and 1C). IL-1RA antibody–positive patients with biopsy-confirmed myocarditis had an early onset of symptoms, which occurred mostly after receipt of the second vaccine dose, and a milder course of myocarditis than patients with biopsy-confirmed myocarditis but without anti–IL-1RA autoantibodies (Tables S1 through S6 and Figs. S1 through S6 in the Supplementary Appendix, available with the full text of this letter at NEJM.org).[1,2]

Figure 1

Autoantibodies Targeting IL-1RA in Myocarditis after SARS-CoV-2 Vaccination.

Blood plasma samples were obtained from 69 patients with suspected myocarditis after receipt of vaccine against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). In 61 patients, endomyocardial biopsy (EMB) was performed, and myocarditis was confirmed on EMB in 40 patients (Panel A). Plasma samples that were obtained from 8 patients with no confirmatory investigation on EMB, from 21 patients in whom the diagnosis of myocarditis was ruled out, and from 40 patients in whom myocarditis was confirmed on EMB were analyzed for antibodies against endogenous interleukin-1 receptor antagonist (IL-1RA) and progranulin by enzyme-linked immunosorbent assay (ELISA). Data are shown sorted according to the age of the study participants (Panel B). OD490 denotes optical density as measured at a wavelength of 490 nm. The frequency of anti–IL-1RA antibodies in plasma samples from patients with vaccine-associated myocarditis was confirmed or ruled out on EMB and was sorted according to age. Control participants were 214 healthy adults who had samples obtained 1 week after receipt of the second dose of SARS-CoV-2 vaccine and 127 patients with myocarditis whose samples were obtained before 2020 (Panel C).

Western blots of native gradient polyacrylamide gel electrophoresis (PAGE) revealed immune-complexed IL-1RA and weakened bands resembling free IL-1RA (16 kDa) in plasma samples seropositive for anti–IL-1RA. In identical samples, isoelectric focusing of IL-1RA revealed a differentially charged IL-1RA isoform (Panel D). Multiple Spearman’s correlation analyses were conducted of IL-1RA plasma levels in anti–IL-1RA–positive patients (left graph) and autoantibody-negative patients (right graph) with the use of troponin T (Trop T; in units per milliliter), creatine kinase (CK and CK-MB; in picograms per milliliter), pro–B-type natriuretic peptide (pro-BNP; in units per milliliter), and CD3+ cells (normal value, <7 per square millimeter) and CD68+ cells (per square millimeter) infiltrating the tissue of the right or left ventricle, respectively, as well as C-reactive protein (CRP; in milligrams per deciliter). Numbers indicate the respective Spearman’s r (*P<0.05, and **P<0.01) (Panel E). IL-1RA plasma levels were determined by ELISA in patients with vaccination-associated myocarditis. Data are shown as violin plots; in each plot, dots indicate individual samples, the solid horizontal line the median, dotted horizontal lines the upper and lower quartiles, and the shaded area the probability density. Data were analyzed by Brown-Forsythe and Welch analysis of variance and Dunnett’s T3 multiple comparisons test (Panel F). Human embryonic kidney IL-1 reporter cells (releasing secreted embryonic alkaline phosphatase on IL-1β signaling) were incubated with tumor necrosis factor α (TNF-α), IL-1β , and IL-1β with anakinra or recombinant human IL-1RA (rec hIL-1RA). Plasma from adult patients with critical coronavirus disease 2019, without and with IL-1RA antibodies, and from patients with myocarditis after SARS-CoV-2 vaccination, without or with IL-1RA antibodies, were added (all plasma in 1:20 dilution). Recombinant anti–IL-1RA antibody and recombinant anti–stomatin-like protein 2 antibody were used as positive and negative controls, respectively (Panel G). 𝙸 bars indicate the standard deviation of the mean. OD650 denotes optical density as measured at a wavelength of 650 nm, and VAM vaccination-associated myocarditis.

IL-1RA antibodies were observed in 2 of 214 vaccinated control participants (1%) and in 2 of 125 participants (2%) who had histologically proven myocarditis that had been diagnosed before the Covid-19 pandemic. Previous data that had been obtained from patients with critical Covid-19 did not support the cross-reactivity of purified IL-1RA antibodies with structural proteins of SARS-CoV-2, including the spike protein,[3] which argues against virus- or vaccine-driven molecular mimicry.

Current evidence points toward a transient hyperphosphorylation of IL-1RA preceding a breakdown of peripheral immune tolerance.[3,4] In Western blots of total plasma protein, antibodies to IL-1RA coincided with weaker bands of free IL-1RA, but prominent immune (IgM or IgG)–complexed protein with an atypical IL-1RA isoform occurred exclusively in patients who were seropositive for anti–IL-1RA antibodies (Figure 1D). This additional IL-1RA isoform was hyperphosphorylated at threonine position 111, which had been observed previously in adult patients with critical Covid-19 and in patients with MIS-C.[3,4] In contrast to our observations in patients with myocarditis after SARS-CoV-2 vaccination, IL-1RA was not hyperphosphorylated in any of the samples that had been obtained from control participants.

At the time of acute myocarditis, the mean (±SD) free IL-1RA plasma level in 15 patients who were seropositive for anti–IL-1RA antibodies was 236±82 pg per milliliter, whereas the level was 1736±312 pg per milliliter in 33 patients without anti–IL-1RA antibodies and 1599±277 pg per milliliter in 21 patients in whom histologic testing ruled out the diagnosis of myocarditis (Figure 1F). IL-1RA plasma levels correlated with markers of cardiac damage (troponin T, creatine kinase, creatine kinase MB, or pro–B-type natriuretic peptide), cardiac-tissue infiltration of CD3+ T cells and CD68+ macrophages, and systemic inflammation (C-reactive protein). There was a negative correlation between markers of cardiac damage and IL-1RA plasma levels in patients with anti–IL-1RA antibodies (Figure 1E). Interleukin-1 signaling reporter assay experiments showed direct impairment of IL-1RA bioactivity after the addition of anti–IL-1RA antibodies from patients’ plasma (Figure 1G).

Our study of SARS-CoV-2 vaccination–associated myocarditis and anti–IL-1RA antibodies should be interpreted within the context that the transiency of hyperphosphorylation (as previously reported in patients with critical Covid-19 or MIS-C[3,4]) and patients’ HLA haplotypes were not known. In our study, neutralizing antibodies against IL-1RA and a hyperphosphorylated IL-1RA isoform were observed in young male patients with biopsy-confirmed myocarditis after the receipt of SARS-CoV-2 mRNA vaccine. These antibodies impaired IL-1RA bioactivity in vitro, were associated with low circulating levels of IL-1RA, and were found in patients with biomarker evidence of cardiac damage and inflammation.