Behind Anti‐MDA5 Antibody: Symbol of A Specific Disease or A Phenomenon?

First reported in 2005, anti-melanoma differentiation-associated gene 5 (MDA5) antibody-positive (anti-MDA5+) dermatomyositis (anti-MDA5+ DM) is thought to be rare and is often complicated by rapidly progressive interstitial lung disease (RP-ILD), resulting in poor prognosis [1]. Anti-MDA5+ DM has mainly been reported in Asian populations, mostly in China and Japan, but has been gaining increased attention worldwide, especially after the COVID-19 pandemic. The crucial problems in anti-MDA5+ DM management are stratifying patients according to their prognosis and exploring optimal therapeutics. This mini-review focuses on prognostic factors, the frontier of therapeutic strategies, and the recent progress in understanding their pathophysiology. Two recent meta-analyses showed similar results for the prognostic factors of anti-MDA5+ DM [2, 3]. The mortality risk associated with anti-MDA5+ DM increases with age. Many patients with anti-MDA5+ DM exhibit little muscle involvement but are characterized by skin and lung involvement. RP-ILD is a unique phenotype of anti-MDA5+ DM with a poor prognosis. However, skin lesions (heliotrope rash, Gottron's sign, skin erosions, and ulcers) were not associated with poor prognosis in anti-MDA5+ DM. Elevated lactate dehydrogenase (LDH), creatinine kinase, ferritin, C-reactive protein, lymphocytopenia, and anti-Ro 52 positivity predict mortality in patients with anti-MDA5+ DM. Studies have also reported the predictive value of several pro-inflammatory cytokines, including interleukin (IL)-6, IL-15, and IL-18 [2, 4]. Anti-MDA5 antibody serves not only as a diagnosis biomarker; quantification of anti-MDA5 antibody levels also appears to be useful for disease activity assessment and predicting treatment response [4]. Several prognostic models have been proposed to predict the mortality of patients with myositis and interstitial lung disease, including the FLAIR and MCK models [1, 5]. Additionally, the CROSS model effectively predicted the incidence of RP-ILD in patients with anti-MDA5+ DM [6]. Factors comprising these models, such as ferritin, LDH, C-reactive protein, and anti-Ro 52-antibody positivity, can be adopted for the mortality prediction of anti-MDA5+ DM. However, other aforementioned factors could also serve as supplements and should further be investigated in prospective studies. Immunosuppressive therapy is the mainstay treatment for MDA5 + DM patients, showing variable efficacy. Triple combination therapy comprising high-dose glucocorticoids, calcineurin inhibitors, and intravenous cyclophosphamide has demonstrated favorable survival rates in anti-MDA5+ DM patients compared with dual therapy (glucocorticoids in combination with calcineurin inhibitors or cyclophosphamide) or step-wise addition of these two immunosuppressants on the basis of high-dose glucocorticoids [7]. The effectiveness of triple combination therapy indicates the need to target lymphocytes with different functions and comprehensively block the inflammatory responses in patients with anti-MDA5+ DM. Therapies targeting autoantibody production or enhancing its clearance, including rituximab and therapeutic plasma exchange (TPE), have also been used in these patients. Several case series have reported clinical improvement in anti-MDA5+ DM patients, mostly with RP-ILD, after administration of rituximab [8]. However, the reported efficacy of rituximab varied in anti-MDA5+ DM between studies. Shirakashi et al. [9] reported that among 13 anti-MDA5+ DM patients who were resistant to the triple combination therapy, five of eight patients who received TPE survived, while untreated five patients died (p = 0.04) at an average of 2.4 months. Since then, TPE has been introduced as an adjuvant treatment [8]. Biological agents targeting inflammatory pathways, including IL-6 receptor antagonists and Janus kinase inhibitors (JAKi), have been used to treat patients with anti-MDA5+ DM, and successful results have been reported occasionally [10]. In a Chinese study, the JAKi protocol, using tofacitinib at a dose of 5 mg twice daily combined with glucocorticoids, significantly improved the survival of early-stage anti-MDA5+ DM patients [11]. However, direct comparisons between triple combination therapy and biological agents are lacking. The outcomes of patients with severe MDA5+ DM are still far from ideal. Effective therapeutic strategies should be explored in prospective trials. Anti-MDA5 dermatomyositis has occasionally been reported in patients with viral infections or malignant tumors. In addition, anti-MDA5 autoantibody positivity is not limited to dermatomyositis [12]. The presence of anti-MDA5 antibodies and clinical symptoms similar to those of anti-MDA5+ DM, including RP-ILD and lung injury, have been reported in severe COVID-19 patients [13]. Recently, we also reported an immune-related adverse event (irAE) case presenting as RP-ILD with anti-MDA5 antibody positivity in a patient with lung cancer, indicating that immune checkpoint inhibitors could be a trigger for the development of anti-MDA5+ RP-ILD [14]. Based on this evidences, autoantibodies directed against MDA5 likely connect different disease domains and endow patients with similar features, especially RP-ILD and poor prognosis. As a cytosolic pattern recognition receptor for viral RNA, MDA5 plays an essential role in anti-viral immune response by activating the IFN pathway. Plasma IFN-α levels were found to be increased in anti-MDA5+ DM patients, compared with patients suffering from other types of DM. Recently, Ye et al. [15] reported that type I IFN signaling is overstimulated in peripheral B cells and T cells obtained from patients with anti-MDA5+ DM, using single-cell RNA sequencing. Moreover, they identified pro-fibrotic response via type I IFN signaling in the affected lung tissue samples [15]. Besides, the IFN signature has been identified in the skin, and muscle of patients with anti-MDA5+ DM [16]. These findings suggest a crucial role of type I IFN pathway in the pathogenesis of anti-MDA5+ DM. This common pathway may explain the similar presentations observed in different clinical scenarios. The pathogenic effect of the anti-MDA5 antibody has not been fully established. A recent study identified lung injury after treating human MDA5 transgenic mice with rabbit anti-human MDA5 polyclonal antibodies [17]. Meanwhile, Ichimura et al. recently demonstrated in a mouse model, that the combination of MDA5-immunization and administration of poly (I:C), a mimicker of viral infection, but not MDA5-immunization alone, resulted in the development of ILD [18]. Therefore, we speculate that anti-MDA5 antibodies could trigger the ILD development in the presence of several factors, such as infections and drugs in genetically susceptible patients. Type I IFN activation in sequence aims at fighting against these triggers but also triggers uncontrolled immune responses, including dysregulated biomarkers and immune cells, resulting in damage to different organs (lung, skin, and muscle), as seen in anti-MDA5+ DM (Figure 1). Studies have shown that a considerable number of patients with anti-MDA5+ DM die during the first 6 months of onset, and survivors could anticipate extended survival, which is supposed to be related to fluctuations in Type I IFN signaling pathway activation [2]. In conclusion, anti-MDA5 antibody positivity may be shared among patients with inflammatory diseases, infectious diseases, and cancer. Anti-MDA5+ autoantibodies could represent overly uncontrolled activation of the IFN pathway. This might expand our understanding of this syndrome and its underlying mechanisms, which will shed light on the lives of patients with anti-MDA5+ DM. All authors contributed to the conception of this article. Huaiya Xie drafted the manuscript. Junping Fan critically revised the manuscript. Luo Wang and Xinlun Tian provided significant revisions to the manuscript. All authors approved the final version of the manuscript and agreed to be accountable for all aspects of the work. The authors have nothing to report. The authors declare no conflicts of interest. The authors confirm that the data supporting the findings of this study are available within the article.