The comparison of nailfold videocapillaroscopy findings between anti-melanoma differentiation-associated gene 5 antibody and anti-aminoacyl tRNA synthetase antibody in patients with dermatomyositis complicated by interstitial lung disease

Dermatomyositis (DM) is frequently complicated by interstitial lung disease (ILD), which increases mortality. This study aims to elucidate the clinical significance of nailfold videocapillaroscopy (NVC) on assessing the disease activity and prognosis of DM-ILD. We compared the NVC findings between anti-melanoma differentiation-associated gene 5 (anti-MDA5) antibody-positive and anti-aminoacyl tRNA synthetase (anti-ARS) antibody-positive patients, the survival and ILD-related death groups, and examined the association of NVC findings with prognostic factors of DM-ILD. The median scores of microhemorrhage and capillary disorganization in the anti-MDA5 antibody-positive group were significantly higher than those in the anti-ARS antibody-positive group (P = 0.012 and 0.044, respectively). In contrast, the median scores of tortuous capillaries in the anti-ARS antibody-positive group were significantly higher than those in the anti-MDA5 antibody-positive group (P = 0.002). The median scores of microhemorrhage was significantly higher in the ILD-related death group than the survival group (P = 0.02). The scores of microhemorrhage, capillary disorganization, and neoangiogenesis correlated with known poor prognosis factors of DM-ILD. Additionally, the scores of microhemorrhage and capillary loss correlated significantly with the total fibrosis scores of chest high-resolution computed tomography. These findings suggest that NVC is a useful tool for assessing the disease activity and prognosis of DM-ILD.

Treatment for patients with DM-ILD. Prednisolone (PDN) (0.5-1.0 mg/kg/day) was administered in 22 of the 27 patients. Cyclosporine (CSA) or tacrolimus (TAC) was used as combination treatment according to the physician's decision. CSA was started at 4 mg/kg/day once a day before breakfast, and the concentration at 2 h after administration was adjusted to 1,500 ng/mL or above. TAC was started at 0.1 mg/kg/day twice a day before breakfast and dinner, and the trough was adjusted to 5-15 ng/mL 21 . Whether additional treatments such as methylprednisolone pulse therapy (MPDN), intravenous pulse cyclophosphamide, or intravenous immunoglobulin were administered was determined by the physician depending on each patient's condition.
Arterial blood gas analysis and pulmonary function test. Arterial blood gas analysis including PaO 2 , PaCO 2 , and AaDO 2 was conducted on admission. Respiratory function was measured by spirometry (SYSTEM21; Minato Medical Science, Osaka, Japan). Vital capacity was determined by the N2 washout method, and diffusion capacity of the lung for carbon monoxide was determined by the single-breath method [22][23][24] . Respiratory function test results are expressed as percentages of the predicted value.
HRCT scoring. HRCT was performed using a 64-detector row CT Aquilon multiscanner (Toshiba Medical Systems Corporation, Tokyo, Japan). Slice thickness was 1.0-1.5 mm every 10 mm, with the scan area including the entire lung. All patients underwent chest HRCT prior to treatment, and images were reviewed independently by 3 observers (SM, TK, and TS) blinded to the patients' clinical information. Inter-observer disagreements were resolved by consensus. Ground-glass opacity (GGO) and fibrosis were both scored to assess HRCT findings as described previously 25 . The lobes of each patient were scored by the same observers, and the average of the three values was used. The scores obtained were summed as the total CT score. Because the right middle lobe GGO score is strongly related to the poor prognosis of DM-ILD 26 , it was considered as a poor prognostic factor. nailfold videocapillaroscopy. NVC was performed using a Dino-lite capillaroscopy device (with Dinocapture 2.0 windows software) at 200 × magnification. Patients were acclimated to a room temperature of 20-23 °C at least 15 min before the exam. Immersion oil was placed on the nailfold beds to improve the image resolution. All images were analyzed by two independent rheumatologists (RW and TS) who were blinded to the patients' clinical diagnosis and disease severity. We took two images of the middle of the nailfold for all fingers of both hands excluding the thumbs.
A semiquantitative rating scale to score NVC findings was adopted as follows according to previous studies 27, 28 ; 0 = no changes; 1 = less than 33% of capillary alterations/reduction, 2 = 33-66% of capillary alterations/
Other laboratory findings, such as clinical symptoms, AaDO 2 , respiratory function test results, and chest HRCT scoring results are listed in Table 1. PDN was used in 22 patients, and the median dose was 45 (30-56) mg/day. CSA were used in 8 patients and the median doses were 200 (131-250) mg/day. TAC were used in 11 patients and the median doses were 4 (3-4) mg/day. MPDN pulse and intravenous immunoglobulin therapy were used in 5 and 8 patients, respectively. Intravenous pulse cyclophosphamide was used in 12 patients, and the median total dose was 4,750 (3,000-6,500) mg. All patients who died were anti-MDA5 antibody positive. Comparison of NVC findings between the survival and death due to ILD groups. NVC findings between the survival group and death due to ILD group can be compared in Table 3. The median microhemor-

Chest HRCT scores
Right middle lobe GGO score  Supplementary Table S2, we also compared the differences in NVC findings of anti-MDA5 antibody-positive DM-ILD between survivors (N = 7) and dead patients (N = 3). The median scores for microhemorrhage, capillary loss, capillary disorganization, and neoangiogenesis tended to be higher in the death due to ILD group than those in the survival group (P < 0.2). However, there was no statistical significance between the two groups.

Correlation between NVC findings and disease activity indicators of DM-ILD. The correlations
between NVC findings and disease activity indicators of DM-ILD are shown in Table 4. The microhemorrhage score correlated significantly with the serum levels of CRP (R = 0.48), ferritin (R = 0.41), AaDO2 (R = 0.45), total GGO scores (R = 0.39), and total fibrosis scores (R = 0.40). The score of giant capillaries correlated significantly with the serum levels of ferritin (R = 0.45) and the right middle lobe GGO scores (R = 0.58). The score of capillary disorganization correlated significantly with ferritin (R = 0.51) and with AaDO 2 (R = 0.66). The neoangiogenesis score correlated significantly with the serum levels of ferritin (R = 0.46), AaDO 2 (R = 0.50), and the total GGO scores (R = 0.65). The score of bushy capillaries correlated significantly with the serum levels of ferritin (R = 0.41), Table 2. Comparison of nailfold videocapillaroscopy findings between anti-MDA5 antibody positive and anti-ARS antibody positive DM-ILD patients. MDA5, anti-melanoma differentiation-associated gene 5; ARS, aminoacyl-tRNA synthetase; DM, dermatomyositis; ILD, interstitial lung disease. The laboratory markers are presented as the median (interquartile range). The P-values were estimated using Wilcoxon rank sum test. *P < 0.05, **P < 0.01.

Discussion
In this study, we investigated the relationship between NVC findings and clinical features in DM-ILD patients. A high microhemorrhage score was observed in the NVC findings in the anti-MDA5 antibody-positive patients and the death due to ILD group, and correlated with the poor prognostic factors of DM-ILD. A high tortuous capillary score was observed in anti-ARS antibody-positive patients. The score of capillary loss was associated with the extent of lung fibrosis. The NVC findings were associated with myositis-specific autoantibodies, severity, and the extent of lung fibrosis in DM-ILD patients.
Microhemorrhages are frequently observed in NVC findings in patients with DM, probably due to the microvasculopathy associated with vascular endothelial injury 29 . Mugii et al. reported that microhemorrhage in NVC findings and are correlated with skin disease activity and considered to be reversible because it decreases after immunosuppressive therapy in DM patients 18,30 . Fiorentino et al. reported that the skin ulcers of anti-MDA5 antibody-positive DM patients had spread to the nail side, inflammatory cell infiltrations were observed around the capillaries in the biopsy tissue, and they exhibited severe vasculopathy 31 . Also, Okiyama et al. reported that Table 3. Comparison of nailfold videocapillaroscopy findings between dead groups and alive groups in DM-ILD. DM, dermatomyositis; ILD, interstitial lung disease. The laboratory markers are presented as the median (interquartile range). The P-values were estimated using Wilcoxon rank sum test. *P < 0.05.  Table 4. Correlation of nailfold videocapillaroscopy findings with disease activity factors of DM-ILD. DM, dermatomyositis; ILD, interstitial lung disease; DL, Diffusing capacity of the lung; HRCT, high-resolution computed tomography; CK, creatine kinase; LD, lactate dehydrogenase; CRP, C-reactive protein; KL-6, Krebs von den Lungen-6; AaDO 2 , alveolar-arterial oxygen difference; FVC, forced vital capacity; DLco, diffusion capacity of the lung for carbon monoxide; GGO, ground-glass opacity. Correlations were evaluated using Spearman's rank correlation coefficient. *P < 0.05, **P < 0.01. www.nature.com/scientificreports/ vascular injury in skin biopsy was observed more often in the anti-MDA5 antibody-positive group than in the anti-ARS antibody-positive groups and anti-transcriptional intermediary factor-1 antibody-positive groups 32 . These reports support the results of the present study, which showed that anti-MDA5 antibody-positive DM patients had high scores of microhemorrhage and capillary disorganization in the NVC findings. In the present study, the poor prognostic factors of DM-ILD, such as serum ferritin, CRP, and KL-6 levels, and chest HRCT findings, correlated with the scores of microhemorrhage and capillary disorganization in the NVC findings. Also, the scores of microhemorrhage were significantly higher in the death due to ILD group than that in the survival group. In our study, all dead patients were anti-MDA5 antibody-positive, which indicated severe vasculopathy in skin compared to anti-ARS antibody-positive groups 33 . Therefore, the score of hemorrhage, which reflects the severity of vasculopathy, in dead patients are significantly higher than those in the survival patients. Therefore, microhemorrhages in NVC findings are valuable markers for the evaluation of disease activity and prognosis in DM-ILD.

Findings
In anti-MDA5 antibody-positive DM-ILD, the median scores of microhemorrhage, tended to be higher in the death due to ILD groups than those in the survival group. This result might suggest that the dead group in anti-MDA5 antibody positive DM-ILD showed severe vasculopathy compared to the survival group. However, there was no significance between two groups due to limitation of small sample size.
Tortuous capillaries in NVC findings represent a pathology of neoangiogenesis resulting from chronic capillary injury 29 . This study showed that the positive correlation between the tortuous capillaries score in the NVC findings and the disease duration of DM-ILD patients. Also, the frequency of CIP was significantly higher in patients with anti-ARS antibody than that in patients with anti-MDA5 antibody. Therefore, patients with anti-ARS antibody had higher scores of tortuous capillaries compared to patients with anti-MDA5 antibody. Tortuous capillaries reflecting the neoangiogenesis could be valuable marker for the evaluation of chronicity and regeneration in DM-ILD.
Capillary loss in NVC findings is observed in the late phase in SSc patients and related to the skin fibrosis progression 16,34 . In the present study, there was a significant positive correlation between the score of capillary loss and the total fibrosis score on chest HRCT findings. A high total fibrosis score in the chest HRCT findings of DM-ILD patients expresses the state of ILD progression and the chronic phase after structural modification of the lung 25 . Thus, a high capillary loss score in the NVC findings is considered to represent the advanced and chronic stage of DM-ILD.
To our knowledge, the relationship between pulmonary pathology and NVC findings has not been examined in DM-ILD patients. A pulmonary biopsy was not obtained from all patients in this study and could not be compared with the NVC findings. However, in DM cases, the state of the capillaries in muscle and skin biopsies has been reported to be associated with NVC findings 19,35 , and microvasculopathy in NVC findings may be associated with lung lesions in DM-ILD patients. This is a retrospective study conducted in a single center with small number cases. To further clarify the association of NVC findings with the pathology of DM-ILD patients, it will be necessary to accumulate data by prospective analysis in a multi-center study.

conclusions
In conclusion, we investigated the relationship between NVC findings and clinical features in DM-ILD patients. The NVC findings were associated with myositis-specific autoantibodies, severity, and the extent of lung fibrosis in DM-ILD patients. NVC is a powerful tool for the prediction of disease activity and prognosis of DM-ILD.