Biological therapy improves myocarditis and disease activity in eosinophilic granulomatosis with polyangiitis patients

Cardiac insuciency is a major cause of mortality in eosinophilic granulomatosis with polyangiitis (EGPA). Despite the dosages-related cardiotoxicity, cyclophosphamide is usually prescribed to induce disease remission in the presence of myocarditis with heart involvement. There is an imperative need of novel medications to eciently control disease activity and spare the use of cyclophosphamide. A retrospective study was carried out in hospitalized EGPA patients from January 1, 2008 to December 31, 2019, focusing on the use of biologics including benralizumab (BEN, anti-IL-5 receptor), mepolizumab (MEP, anti-IL-5), omalizumab (OMA, anti-IgE) and rituximab (RTX, anti-CD20). × regimen and 24-hour Holter evaluate the therapeutic responses in involvement. After biological there were improved cardiac dysfunction, lower eosinophil and clinical remission (4 complete, activity or relapsing disease. Two obtained MEP under 100 mg quadri-weekly subcutaneous injection (cases no. 1 and 13) due to disease relapse with a more than 90% inhibition (92.9 ± 1.3) of the baseline eosinophil counts. Case no. 16 acquired OMA under 150 mg bi-weekly subcutaneous injection due to relapsing disease at the age of 12 [20], and Ben with 30 mg quadri-weekly subcutaneous injection at the age of 16 with a 100% inhibition of the baseline eosinophil counts due to disease relapse [12]. For those receiving biologics, there were complete remission in 5 (cases no. 1, 2, 3, 4 and 16) and partial remission in 4 (cases no. 5, 6, 7 and 13). In particular, CS and immunosuppressants were not prescribed in cases no. 1 and 16 after biologic therapy.

In the existence of major organ damage with lethal manifestations like cardiac insu ciency, in addition to corticosteroids (CS), cyclophosphamide (CYC) is usually prescribed to induce disease remission [6]. Nevertheless, the majority of deaths in EGPA are caused by disease activity with heart involvement as a leading cause in spite of under combined CS and CYC therapy [4,6,7]. Furthermore, given the signi cant toxicity of CYC, it is generally acknowledged not to exceed 10 to 15 gram of exposure in EGPA patients due to the dosages-related cardiotoxicity [7]. Indeed, there is an imperative need of novel therapeutics to e ciently control disease activity and spare the use of CS and CYC [8].
Notably, eosinophils play a central role in the EGPA pathogenesis with eosinophilia of blood and tissue as the disease hallmark, and IL-5, a Th2 cytokine, is recognized as the key mediator in the development and maintenance of eosinophilia [1,8,9]. Through the autoantigen presentation and costimulatory signaling to Th2 cells, B cells can participate in the EGPA pathogenesis by inducing the release of IL-5, resulting in activation, maturation, survival and recruitment of eosinophils [1,8,10]. Current evidence supports the use of mepolizumab (MEP), an IL-5 monoclonal antibody (mAb), for induction treatment of refractory or relapsing EGPA with greater eosinophilia [1,11]. Despite the ongoing clinical trials exploring the e cacy of benralizumab (BEN), an IL-5 receptor mAb, in EGPA, this biologics has acquired the orphan drug designation from the US regulator [1,12]. Therapeutic responses to rituximab (RTX), a B celldepleting mAb, have been observed in EGPA, and it has been recommended at induction for refractory activity or relapsing disease with a potential bene t in ANCA negativity [13]. Interestingly, MEP has been suggested to treat eosinophilic myocarditis in hypereosinophilic syndrome (HES) [14], and therapeutic effects of RTX on autoimmune myocarditis have been observed in microscopic polyangiitis and systemic lupus erythematosus [15,16].
To identify myocarditis in EGPA, instead of investigating stable victims without admission, we analyzed hospitalized patients with disease activity and focused on the use of biologics in this retrospective study.

Patients enrollment
Under the approval of the Institutional Review Board, patients admitted to Departments of Internal Medicine and Pediatrics and ful lling 1990 American College of Rheumatology (ACR) criteria for the EGPA classi cation [17], were analyzed from January 1, 2008 to December 31, 2019. Myocarditis was diagnosed according to the following criteria: (1) presenting symptoms consistent with heart failure, (2) raised concentrations of cardiac biomarkers, and (3) new or worsening changes including impaired left ventricle ejection fraction (LVEF, mild 46 to 55%, moderate 30 to 45%, severe below 30%) or plus wall motion abnormality on transthoracic echocardiography (ECG) or cardiac magnetic resonance imaging (cMRI) [15,16,18]. The additional endocardial involvement, i.e., endomyocarditis, was detected by cMRI. Cardiac rhythm was examined by 24-hour Holter monitor. Exclusion criterion was coronary artery disease with characterized ndings in coronary angiography or cMRI, viral myocarditis with a con rmed history of infection, or preexisting heart diseases under medical therapy [15,16].

Data Collection
Demographical, clinical, laboratory, imaging and pathological data were analyzed, including age/sex, EGPA manifestations, disease activity (Birmingham Vasculitis Activity Score, BVAS) [19], prognosis assessment [5], myocarditis symptoms [15,16], New York Heart Association Functional Classi cation (NYHAFC), ANCA titers (immuno uorescence, enzyme-linked immunosorbent assay), eosinophil counts, C-reactive protein (CRP) values, IgE levels, circulating B-cell numbers (CD19-positive lymphocytes), cardiac biomarker concentrations (cardiac troponin I, creatine kinase-MB, N-terminal pro-brain natriuretic peptide), cardiac image and rhythm ndings, and pathological results. There was a comprehensive review in medication pro les including CS, immunosuppressive agents, biologics and cardiac medications with antiarrhythmic drugs for cardiac dysrhythmia and cardiac supportive agents (CSA) (angiotensin converting enzyme inhibitor, angiotensin-receptor blocker, β blocker, diuretic, inotrope) for heart failure. A complete remission was de ned as the absence of disease activity corresponding to zero BVAS, and a partial remission was a reduction of no less than 50% in BVAS as compared with the baseline scores.

Statistical analysis
Data was expressed as the mean and standard deviation. Numerical data between two groups were compared by the Mann-Whitney test. BVAS, CRP levels and eosinophil counts before and after the biologic therapy were calculated by the Wilcoxon signed rank test. P value less than 0.05 was considered signi cant in this study.

Characteristics of admitted patients
In Table 1, 16 admitted patients ful lled 5 or 6 items (5.4 ± 0.5) of the ACR classi cation criteria [17], and had histopathological ndings of tissue eosinophilia or plus vasculitis of small-to medium-sized vessels (Fig. 1). There were 8 females aged 10 to 70 years (40.4 ± 15.5), with positive ANCA against myeloperoxidase in 5 (31%), initial BVAS 16 to 39 (26.8 ± 6.9) and FFS 1 or 2 (1.4 ± 0.5). Laboratory parameters at the disease onset were eosinophil percentages 21 to 79% (46.6 ± 15.2) with total eosinophil counts 2,314 to 26,781/µL (11,108 ± 7,060), CRP values 19.1 to 183.3 mg/L (79.9 ± 51.6) and IgE levels 123 to 4,000 IU/mL (1,041 ± 1,002). Despite the challenge in differentiating idiopathic HES from EGPA due to eosinophilic tissue in ltration in both disorders, the presence of asthma is a characteristic diagnostic feature of EGPA [6,7]. All enrolled patients had recurrent asthmatic attacks. Involvement of lung parenchymal, peripheral nerve system, skin, heart, sinus, joint, muscle, kidney, gastrointestinal tract and central nervous system was identi ed in 14, 14, 13, 10, 9, 7, 6, 4, 3 and one, respectively. Patients with cardiac manifestations had higher initial eosinophil counts than those without heart involvement (12,731 ± 6,346 versus 6,737 ± 4,439/µL, P = 0.056). In this series, two (cases no. 10 and 15) expired due to disease activity, and fourteen survived with complete remission in 5 and partial remission in 9. For the biologics use, 6 received RTX under 375 mg/m 2 weekly intravenous infusion (cases no. 2 to 7) due to refractory activity or relapsing disease. Two obtained MEP under 100 mg quadri-weekly subcutaneous injection (cases no. 1 and 13) due to disease relapse with a more than 90% inhibition (92.9 ± 1.3) of the baseline eosinophil counts. Case no. 16 acquired OMA under 150 mg bi-weekly subcutaneous injection due to relapsing disease at the age of 12 [20], and Ben with 30 mg quadri-weekly subcutaneous injection at the age of 16 with a 100% inhibition of the baseline eosinophil counts due to disease relapse [12]. For those receiving biologics, there were complete remission in 5 (cases no. 1, 2, 3, 4 and 16) and partial remission in 4 (cases no. 5, 6, 7 and 13). In particular, CS and immunosuppressants were not prescribed in cases no. 1 and 16 after biologic therapy.
Cases no. 4, 5, 8 and 10 had concurrent pericardial effusion, consistent with the diagnosis of myopericarditis [18], and cases no. 1, 4, 5 and 7 had coexistent endocarditis, an ominous manifestation in EGPA associated with overt heart failure [21]. Patients with additional pericardial or endocardial involvement, indicative of diffuse heart involvement, had lower LVEF than those without such a presentation (for pericarditis, 38.8 ± 7.6% versus 44.3 ± 10.1%; for endocarditis, 38.5 ± 11.2% versus 44.5 ± 7.6%). For myocarditis-related cardiac arrhythmia, 9 had sinus tachycardia with additional ventricular extrasystoles in 3 (no. 3, 4 and 10) and atrial extrasystoles in 2 (no. 1 and 10). In particular, one had sinus bradycardia complicated with sinus pause (no. 2). All received antiarrhythmic drugs and CSA for their underlying cardiac dysfunction. Table 2 shows clinical, laboratory and medication pro les, and MPA-or RTX-related therapeutic indication and regimen in 7 patients receiving biologic therapy.  Besides lower CRP levels (25.2 ± 14.4 to 2.6 ± 1.8 mg/L, P = 0.016) after biologic therapy, there was a decrease in BVAS (20.1 ± 5.2 to 1.9 ± 2.5, P = 0.016) with a complete remission in 4 patients and a partial remission in 3. Peripheral eosinophil counts were reduced from 989 ± 174 to 200 ± 132/µL (P = 0.016) with a 79.8 ± 12.4% inhibition of the baseline values. Before biological therapy, all received the use of CS and CYC at induction, and azathioprine was prescribed at maintenance before the relapsing disease in cases no. 1, 2, 3 and 7. The accumulated CYC dosages were beyond 15 gram in cases no. 2 and 3, and 10 gram in cases no. 4, 5 and 6. During biological therapeutic period, all received cardiac medications. After biological therapy, one received azathioprine alone, 2 obtained low-dose CS (5 mg/day prednisolone) alone, and 3 acquired both drugs. For the CSA prescription, case no. 2 received an angiotensin-receptor blocker and cases no. 3, 4, 6 and 7 obtained an angiotensin converting enzyme inhibitor. Table 3 demonstrates the myocarditis-related clinical and imaging ndings before and after biological therapy. At the onset of myocarditis, all had clinical symptoms with NYHAFC II in 3 and III in 4, cardiac dysrhythmia, elevated cardiac biomarker concentrations, and lower LVEF with mild impairment in 3 and moderate impairment in 4 (31 to 55%, 40.9 ± 10.5%). LV dilation or global hypokinesia were found in 6 patients (cases no. 2 to 7). Myocardial edema was not identi ed in case no. 5 due to an initial cMRI performed after 2 RTX therapeutic courses. Mid-wall myocardium delayed gadolinium enhancement (DGE) were detected in all. Furthermore, case no. 4 and 5 had pericardial effusion (myopericarditis) and cases no. 1, 4, 5 and 7 had endocardium DGE (endomyocarditis). In addition to improved NYHAFC with normalized LVEF, biomarker concentrations and cardiac rhythm, case no. 1 had reduced myocardial edema and myocardium DGE after MEP induction. Furthermore, there was evidently lessened endocarditis as shown in her follow-up cMRI (Fig. 2).

E cacy In Cardiac Manifestations
Cases no. 2 to 7 had improved NYHAFC, normalized biomarker concentrations, cardiac rhythm, LVEF and LV size/motion, resolved myocardial edema, and reduced myocardium DGE after RTX therapy. Case no. 4 had worsening endocardium DGE in spite of RTX induction; however, reduced endocardial involvement was found after maintenance therapy with another two therapeutic courses. Case no. 7 had unresolved endocardium DGE after RTX induction treatment. Serial cMRI in cases no. 3 and 4 were shown in Figs. 3 and 4, respectively.
Biologics was not prescribed in another 3 patients with cardiac presentations at disease onset, myocarditis in case no. 9 and myopericarditis in cases 8 and 10. They received combined CS and CYC treatment at induction. Improved myocardial dysfunction and a partial BVAS response were observed in cases no. 8 and 9. Nevertheless, case no. 10 with moderately impaired LVEF had persistent cardiac dysfunction without disease remission, and further succumbed to the EGPA activity.

Discussion
Cardiac involvement with heart failure in EGPA is a major cause of early death and a poor long-term prognostic factor [4,22]. Myocarditis usually presents as non-ischemic cardiomyopathy with heart failure and arrhythmia [18]. cMRI can serve as a non-invasive tool for evaluating the myocardium and the endocardium, assessing the extent of heart involvement and helping the evaluation of therapeutic responses. Combined T2-weighted and post-gadolinium T1weighted cMRI images provide the best diagnostic sensitivity and speci city [18,23]. T2-weighted images are allowed to detect myocardial edema, whereas T1-weighted DGE can identify myocardial and endocardial brosis in addition to acute in ammation. In this study, the diagnosis of myocarditis and endomyocarditis were based on both T2-and T1-weighted images, and the therapeutic responses were evaluated by serial follow-up of cMRI. Six patients had myocardial edema before biological therapy, and 7 had mid-wall myocardium DGE and 4 had additional endocardium DGE before or during biological therapy.
Notably, there were reduced myocardial DGE and lessened endocardium DGE after MEP therapy, and resolved myocardial edema and reduced myocardium DGE after RTX treatment.
A randomized control trial of MEP therapy in EGPA with a 300 mg quadri-weekly × 13 regimen for 52 weeks, has demonstrated the e cacy at induction for relapsing and refractory disease as well as for CS-dependent victims [24]. As comparing MEP with placebo, there was a larger proportion of patients in remission, with lower relapse rates and bene ting from CS dose tapering. Nevertheless, owing to no speci c dose evaluation in that study, it remains to be determined whether 300 mg is superior to 100 mg dosage for EGPA therapy [8]. Subsequent trials with a regimen of 100 mg quadri-weekly have been carried out in EGPA with relapsing disease [25,26]. In particular, a clinic cohort with disease relapse under the long-term CS use received low-dose MEP therapy for 16 weeks, resulting in clinical improvement with weaning off CS in all enrolled patients [25]. In this study, a similar 100 mg dosage was prescribed in a EGPA victim for 52 weeks at induction for relapsing disease (case no. 1), leading to a complete remission with sparing the use of CS. Currently, despite the established e cacy of MEP, optimal regimens in treating EGPA remain to be a research focus.
Owing to non-inferiority to CYC, RTX with a 375 mg/m 2 weekly × 4 regimen can be prescribed in severe AAV at induction as the rst-line therapy [2]. A 24-month RTX maintenance regimen with 1 gram every 6 months has shown a lower relapse rate in refractory or relapsing AAV [27]. Another RTX maintaining therapy with a regimen of 500 mg fortnightly at 6, 12 and 18 months, has also demonstrated a sustained remission with overall survival superior to the azathioprinebased protocol in newly diagnosed or relapsing AAV [28]. Despite the exclusion of EGPA from AAV in above studies, therapeutic effects of RTX at induction in refractory or relapsing EGPA have been shown in 2 larger trials with 1 gram fortnightly and 375 mg/m 2 weekly × 4 regimens [20,29]. Furthermore, a regimen with 1 gram fortnightly every 6 months has been observed to be effective as maintenance treatment [29]. A recent investigation reveals a bene cent outcome in the RTX therapy with multiple therapeutic courses (mean 4.6 courses) for relapsing disease and remission maintenance in EGPA patients [30]. In this study, therapeutic bene ts were observed in 5 patients receiving multiple therapeutic courses (mean 3.0 courses) for refractory activity or relapsing disease with a 375 mg/m 2 weekly × 4 regimen at induction and plus a yearly maintenance schedule.
Up to 95% of AAV patients other than EGPA have the presence of ANCA [6,7]. Nevertheless, this autoantibody is only identi ed in around 40% of EGPA patients, implicating the existence of different disease subsets with distinct pathogenic mechanisms according to the ANCA status [31], ANCA-negative EGPA patients are less likely to have typical features of other AAV disorders, but more susceptible to cardiac manifestations [2,8]. EGPA patients have elevated circulating Th17 frequencies and IgG4 levels correlating with their disease activity [32,33]. RTX therapy has been observed to improve disease activity through reducing synovial Th17 numbers in rheumatoid arthritis, and induce therapeutic responses by lowering serum IgG4 levels in IgG4-related disease [34,35].
Moreover, for AAV not including EGPA, disease remission and relapse-free survival after RTX therapy have been observed to be irrelevant to the presence of ANCA [36].
EGPA patients with heart involvement have been shown to have higher eosinophil numbers in peripheral blood than those without cardiac manifestations [37], as also demonstrated in our patients. In addition, there were signi cantly lower circulating eosinophil counts in this study after biological therapy. In addition to the presence of vasculitis lesions, persistent eosinophilia can cause myocardium damage, typically in the form of eosinophilic myocarditis with EGPA as an underlying cause [14]. Furthermore, the endocardium and the underlying myocardium are involved in the eosinophilic endomyocarditis, the most characteristic cardiac abnormality in HES [38]. By using the cMRI survey, 13 (27%) in a 49-patient EGPA cohort were documented to have this speci c form of heat involvement [21], a similar occurrence (4 of 16 patients, 25%) in the present series. In eosinophilic myocarditis-related cardiac manifestations, heart injury is due to a direct eosinophil-mediated cytotoxicity, eosinophil-degranulation products released from eosinophils and the recruitment of in ammatory leukocytes by eosinophil-derived cytokines/chemokines [6,11]. Interestingly, stabilization of cardiac dysfunction in eosinophilic myopericarditis with severely impaired LVEF has been observed in an asthmatic patient receiving MEP therapy, 100 mg quadri-weekly for 28 weeks [39]. In this study, there were signi cantly reduced peripheral eosinophil counts to a more than 90% inhibition after anti-IL5 therapy (cases no. 1 and 13) [40], and normalized myocardial contractility with signi cantly reduced endocarditis in EGPA endomyocarditis under the same therapeutic dosages for 52 weeks (case no. 1). Moreover, we demonstrated improved cardiac dysfunction after RTX therapy in 5 ANCA-negative EGPA patients with myocardial involvement. The therapeutic e cacy of RTX in EGPA myocarditis with ANCA negativity appears to include an action mechanism by suppressing IL-5-mediated eosinophilia through the depletion of B cells [1,10].

Conclusions
In this monocentric retrospective study, we observed improved cardiac dysfunction and disease activity after biologic therapy in EGPA patients with myocarditis.