Treatment of Interstitial Lung Disease In Anti-MDA5-Positive Dermatomyositis: A Retrospective Study of 87 Patients

Objectives: The prognosis of anti-melanoma differentiation-associated gene (MDA5) antibody-positive dermatomyositis (DM)-interstitial lung disease (ILD) (anti-MDA5 -positive DM-ILD) is often poor, especially in rapidly progressive interstitial lung disease (RPILD). So far there is no established therapy. This study evaluated the ecacy and safety of pharmacological treatments for anti-MDA5-positive DM-ILD. Methods: This retrospective study comprised 87 anti-MDA5-positive DM-ILD patients. We reviewed the clinical characteristics, laboratory ndings, lung function treatments, and outcomes of the 87 patients. Cox regression analysis was used to identify predictors of 6-month survival. The association between different combined immunosuppressive regimen and 6-month survival were evaluated. Results: High level of CYFRA21-1 and low PaO2/FiO2 ratio were associated with poor prognosis. Corticosteroid (CS) alone treatment group with higher CYFRA21-1 and lower PaO2/FiO2 ratio showed worse 6-month survival than the combination of CS with immunosuppressants group (p<0.01). In addition, tacrolimus/cyclosporine-treated anti-MDA5-positive DM Non-RPILD exhibited a better survival, comparing with tacrolimus/cyclosporine combined intravenous cyclophosphamide (IVCY)-treated patients (p<0.05). Conclusion: Addition of immunosuppressants to CS, were associated with better 6-month survival in anti-MDA5-positive DM-ILD. The triple regimen (CS, tacrolimus/cyclosporine and IVCY) was not superior to duple one (CS, tacrolimus/cyclosporine) in anti-MDA5-positive DM Non-RPILD.


Introduction
Dermatomyositis is a kind of autoimmune in ammatory disease, that characterized by muscle weakness, skin lesions, arthralgia and internal organs damaged, including heart, lungs [1]. Interstitial lung disease (ILD) is a severe complication of DM patients that is associated with poor prognosis of DM patients, especially rapidly progressive interstitial lung disease (RPILD). Myositis speci c antibodies (MSAs) are observed in 30-40% myositis/DM patients which are reported to be associated with different phenotypes of PM patients in terms of distinct clinical manifestations [2]. Among the MSAs, anti-melanoma differentiation-associated gene (MDA5) auto-antibodies, are recently reported to be closely associated with the development of RPILD, especially in patients with typical skin lesions, which is one of the most prognostic factors in DM-ILD patients [3][4][5][6][7]. RPILD has been characterized by accelerated progressive deterioration of respiratory symptom and pulmonary function in three months, that is an important cause of substantial mortality [8,9]. In previous study, we have observed that the serum CYFRA21-1 level was higher in anti-MDA5-positive DM-ILD, especially in RPILD patients, that was a prognostic factor and also associated with disease prognosis [10]. Evidence for the effectiveness of pharmacological treatment for anti-MDA5-positive DM-ILD is limited. Recently, the effective treatments mainly include corticosteroid (CS), calcineurin inhibitors, intravenous cyclophosphamide (IVCY) and plasmapheresis, etc. The conventional treatment including CS alone or a combination of CS and immunosuppressants, is applied to 28-66% of Asia patients with anti-MDA5-positive DM-ILD [11][12][13]. In clinical practice, CS is given at the beginning of diagnosis of anti-MDA5-positive DM-ILD, often in high doses, to most patients who demonstrate RPILD. Calcineurin inhibitors (tacrolimus or cyclosporine) are usually administered in combined with CS. In addition, a combined immunosuppressive therapy including high-dose CS, calcineurin inhibitors, and IVCY might be e cacious [14,15].
Calcineurin inhibitors are the drug with strong immunosuppressive effect, that can down-regulate the production of cytokines in T lymphocytes by inhibition of calcineurin activation, and then suppress the function of T lymphocytes [16]. As a potent immunosuppressive agent, calcineurin inhibitors had been widely used for autoimmune diseases including DM, rheumatoid arthritis (RA) etc. Cyclophosphamide (CTX) as a kind of common immunosuppressive drugs, mainly inhibits immune responses mediated by T and B lymphocytes, and then reduces the in ammatory response induced by a variety of adhesion molecules, chemokines and cytokines. IVCY is given for treatment of progressive interstitial pneumonia in patients with DM, that had a good therapeutic effect [17]. Although CS, calcineurin inhibitors, IVCY, etc are very important for the treatment of anti-MDA5-positive DM-ILD, overtreatment would increase the risk of opportunistic infections to accelerate the death of patients. Few studies investigate the associations of various pharmacological treatments with survival in anti-MDA5-positive DM-ILD patients in China. Therefore, we conducted this retrospective study in an interstitial lung disease center, to investigate the relationship of immunosuppressive treatment with 6-months survival in anti-MDA5-positive DM patients.

Diagnosis of ILD
ILD was diagnosed according to respiratory symptoms, physical examinations, lung function tests and chest high-resolution CT (HRCT) ndings. HRCT was de ned as organizing pneumonia (OP) pattern, nonspeci c interstitial pneumonia (NSIP) pattern and others [19], that were reviewed by the chest radiologist. The patients had a follow-up period for 6 months.

Data collection
Clinical data were obtained from the electronic medical records to determine the characteristics of anti-MDA5-positive DM-ILD patients. We also collected information of lung function and laboratory tests at initial diagnosis. Written informed consents were obtained from all patients prior to the study, and all participants gave the permission for the use of their serum for research purposes. This study was conducted in accordance with the principles of the Declaration of Helsinki (1989) and approved by the Ethics Committee at Nanjing Drum Tower Hospital.

Therapeutic regimen of anti-MDA-positive DM-ILD
All patients were treated with CS therapy (prednisolone initially administered 500 mg/day for 3 days or 1 mg/kg/day). CS dose was tapered during treatment courses (20% decreased every 4 weeks when the dose was higher than 30mg daily, and 8 weeks when it was lower than 30mg daily). Tacrolimus(1-4mg/day) or cyclosporine(100-300mg/day) was combined with CS. Tacrolimus or cyclosporine was adjusted to maintain blood concentration within the range of 10-12 ng/mL or 100-150 ng/mL. CTX was started intravenously at a dose of 400 mg biweekly. IVCY was extended to 400 mg every 4 weeks after the six times.

Statistical analysis
Descriptive statistics for the clinical characteristics are expressed as counts (percentage) and median range. Continuous variables were compared between different groups using the Mann-Whitney U test. Univariate and multivariate Cox regression were performed to establish predictors of survival in anti-MDA5 positive DM-ILD. The cumulative survival rates were calculated by the Kaplan-Meier test. The logrank test was also used to compare the survival rates in each group. The statistical analyses were performed using SPSS for Windows version 19 (IBM Corp. GraphPad Prism Version 6(GraphPad Software, San Diego, CA, USA). A p value < 0.05 was considered statistically signi cant. fever, and only a few patients (23.0%) had muscle weakness. Laboratory ndings showed that lactate dehydrogenase (LDH), c-reactive protein (CRP), and Cytokeratin 19 fragment (CYFRA21-1) levels were elevated and median PaO2/FiO2 ratio was low. The HRCT imaging showed OP/OP-NSIP pattern (78.2%) and NSIP pattern (21.8%). Median forced vital capacity (FVC% predicted) and diffusing capacity of the lung for carbon monoxide (DLCO% predicted) were 61.2% and 55.5% (Table 1). Table 1 Characteristics

Survival analyses according to different treatment
Kaplan-Meier survival method were used to assess the prognosis of patients underwent different treatment. 6-month survival of CS alone-treated was signi cantly lower than those treated with CS combined immunosuppressors (p < 0.01; Fig. 2A). Compared with patients treated with CS combined tacrolimus/cyclosporine and IVCY, survival at 6 months was signi cantly better in the patients treated with CS combined tacrolimus/cyclosporine (p < 0.01; Fig. 2B).    Most of the patients died during the rst 6 months from respiratory failure caused by RPILD. Figure 3 showed Kaplan-Meier survival curves for the tacrolimus/cyclosporine-treated and tacrolimus/cyclosporine combined IVCY-treated in RPILD and Non-RPILD patients. Survival at 6 months in Non-RPILD patients was signi cantly better in the tacrolimus/cyclosporine-treated group than in the tacrolimus/cyclosporine combined IVCY-treated group (p < 0.05; Fig. 3A). No difference was observed in 6 months survival of RPILD patients between two groups (p = 0.138; Fig. 3B). Table 5 showed the detailed features of anti-MDA5-positive DM-Non-RPILD patients. ESR and CYFRA21-1 levels were obviously higher in those receiving triple intensive immunosuppressive therapy. Table 6 showed the characteristics of anti-MDA5-positive DM-RPILD patients in the tacrolimus/cyclosporinetreated and tacrolimus/cyclosporine combined IVCY-treated group. There was no signi cant difference between two groups. (Table 5,6)

Discussion
Although a number of the combined immunosuppressive regimen have been used for anti-MDA5-positive DM-ILD, the e cacy of such pharmacological treatments have not been con rmed in large prospective studies. This study evaluated the associations of 6-month survival with CS alone or use of immunomodulatory therapy (tacrolimus/cyclosporine, IVCY), which are administered in combination with CS in Chinese anti-MDA5-positive DM-ILD patients. The present study clearly indicated that early combined immunosuppressive regimen was better than CS alone for improving 6-month survival of anti-MDA5-positive DM-ILD patients. This nding was consistent with the prior report that in DM-ILD patients, the early combination of CS with immunosuppressants obtained better outcome, compared the other patients who were administered with CS alone [20,21]. In this study, the patients treated with CS alonewhich was correlated with a higher CYFRA21-1 level -leaded to worse prognosis.
It is well known that DM associated ILD can be divided into RPILD and Non-RPILD based on its pathological progression [22,23]. The prognosis of a subset of patients with RPILD complicating anti-MDA5-positive DM, is extremely poor, compared with Non-RPILD [24][25][26][27]. In clinical practice, intensi ed immunomodulators including CS pulsed doses in combination with tacrolimus/cyclosporine and/or IVCY, are sometimes used in DM-ILD patients, especially in RPILD [28,29].
Tacrolimus/cyclosporine as calcineurin inhibitor suppresses expression of interleukin (IL)-2 to reduce the activation of T cells [30]. T cells and alveolar macrophages play an important role in anti-MDA5-positive DM-ILD pathogenesis [10,31]. CTX mainly suppresses the function of B cells for treatment of DM-ILD [32]. Thus, a combined therapy of various immunomodulators can signi cantly increase the risk of infection. Although intensi ed immunosuppressive therapy is considered as necessary for RPILD complicating DM, the evidence of bene t is not conclusive. A previous study showed that early commencement of a regimen therapy including CS, cyclosporine and IVCY may be more e cacious for those patients with RPILD [33]. Early treatment of cyclosporine should be effective to control disease progression of HRCT and improve the survival in DM-ILD [21]. A previous study demonstrated that intensi ed immunosuppressive treatment with CS, tacrolimus, and IVCY improve survival of anti-MDA5positive DM/CADM-ILD patients with well-tolerated adverse events [27]. However, our results indicated that anti-MDA5-positive DM-RPILD patients couldn't obtain better e cacious with a combined regimen (CS, tacrolimus/cyclosporine and IVCY). There was no difference at 6-month survival between treatment with tacrolimus/cyclosporine and combination therapy with tacrolimus/cyclosporine and IVCY in anti-MDA5-positive DM-RPILD patients. In contrast, combination therapy with tacrolimus/cyclosporine and CS was shown to be superior to the combination of tacrolimus/cyclosporine, IVCY and CS in Non-RPILD patients. Non-RPILD patients administered with tacrolimus/cyclosporine, which was associated with a signi cantly lower ESR and CYFRA21-1 levels, -had better prognosis. Furthermore, such patients using single agent such as tacrolimus or cyclosporine, in addition CS, could reduce long-term adverse reactions and obtained better outcome. Thus triple intensive immunosuppressive therapy was not necessary in anti-MDA5-positive DM Non-RPILD in our study.
Overall, we assessed the relationship between intensive immunosuppressive regimen and 6-months survival in anti-MDA5 positive DM-ILD patients. However, there were several limitations in our study. First, other therapies such as plasmapheresis could not be applied generally in subjects due to rarity of plasma. Second, our ability to perform serial analysis of immunosuppressive regimen was limited due to the small study cohort, which was understandable in view of the rarity of anti-MDA5 positive DM-ILD. Third, this was a retrospective single-center study, so it was uncertain whether or not our results could be Not applicable.

Consent for publication:
All author give consent to publish.
A vailability of data and material: The data used to support the ndings of this study are available from the corresponding author upon request Competing interests: The authors have declared no con icts of interest.

Funding
Not applicable.

Authors' contributions:
Yonglong Xiao conceived and designed the study. Xianhua Gui, ShenyunShi and Tingting Zhao collected and analyzed the data. Yuying Qiu and Min Yu contributed to analysis tools. Xianhua Gui wrote the paper. Xiaoyan Xin was a radiologist. The diagnosis of ILD patterns of interstitial pneumonia was based on radiological assessment of the chest HRCT results in the manuscript.
Miao Ma, Jingjing Ding, Lulu Chen, Xiaohua Qiu, Yingwei zhang, Min Cao and Mei Huang were responsible for the data collection in the manuscript.
Mengshu Cao, JinghongDai and Hourong Cai assisted with revising our poor language in the manuscript carefully.
All authors reviewed the manuscript critically and agreed upon publication.   Kaplan-Meier survival curves for the tacrolimus/cyclosporine-treated and tacrolimus/cyclosporine combined IVCY-treated in RPILD and Non-RPILD patients. Survival at 6 months in Non-RPILD patients was signi cantly better in the tacrolimus/cyclosporine-treated group than in the tacrolimus/cyclosporine