Association between nail-fold capillary findings and disease activity in dermatomyositis

Abstract

Objective. Although findings of nail-fold capillary changes and reduced red blood cell velocity in SSc patients are well established, studies in adult-onset DM patients are scarce. Our objective was to assess the changes and red blood cell velocity in finger nail-fold capillaries using nail-fold video capillaroscopy (NVC) in patients with adult-onset DM.

Methods. This study included 50 patients with adult-onset DM and 20 healthy subjects. A semi-quantitative rating scale was used to score capillaroscopy changes. Red blood cell velocity was evaluated using frame-to-frame determination of the position of capillary plasma gaps.

Results. Thirty-seven (74%) patients showed the scleroderma NVC pattern. Patients with the scleroderma pattern exhibited elevated serum creatine kinase levels more frequently and increased visual analogue scale of muscle disease activity. Scores of loss of capillaries were associated with muscle and global disease activity, whereas scores of haemorrhages were associated with skin disease activity. However, NVC findings were not significantly associated with lung involvement. The scores of irregularly enlarged capillaries, haemorrhages and loss of capillaries were reduced after stabilization of disease activity by treatment. The mean red blood cell velocity was not significantly reduced in DM patients compared with healthy controls and was not changed by treatment.

Conclusion. Our results suggest that changes in nail-fold capillaries reflect disease activity in DM. Furthermore, the differences found in red blood cell velocity may reflect somewhat distinct microcirculation injuries in DM and SSc.

Capillaroscopy, Dermatomyositis, Disease activity, Red blood cell velocity, Systemic sclerosis

Introduction

DM is an autoimmune connective tissue disease in which characteristic patterns of inflammatory injury occur in striated muscle. PM is a similar disease, but DM can be distinguished from PM via the presence of cutaneous features such as heliotrope rash or Gottron’s papule/sign. Interstitial pneumonia and internal malignancy are examples of organ complications that affect the prognosis of DM. Pulmonary involvement is frequently seen in patients with anti-aminoacyl tRNA synthetase (ARS) antibodies [1] or anti-clinically amyopathic DM (CADM)-140 antibody [2]. In contrast, a recent study has reported that autoantibodies reactive with 155 kDa (and 140 kDa) nuclear proteins (anti-p155 antibody/anti-155/140 antibody) are associated with malignancy in patients with DM [3, 4]. DM patients with anti-Mi-2 antibodies typically present with relatively mild symptoms and have a good prognosis [5].

A vast amount of well-documented data exist regarding nail-fold capillaroscopy and nail-fold video capillaroscopy (NVC) in the diagnosis and follow-up of microvascular damage in patients with RP or scleroderma spectrum disorder (SSD), SSc and its related diseases [6–11]. Furthermore, NVC changes are just as prevalent and prominent in DM as in SSD [12, 13]. However, such findings are not frequently detected in patients with other connective tissue diseases. Video recordings of blood capillary flow also allow measurement of capillary red blood cell velocity. We recently reported that red blood cell velocity is remarkably reduced in SSc patients compared with healthy individuals [14]. Additionally, reduced peripheral blood flow has been demonstrated using laser-Doppler perfusion imaging [15, 16].

Despite numerous reports regarding NVC findings in SSc, there are relatively few NVC examinations in DM, and this is especially true for adult-onset DM [17–20] compared with JDM [21–23]. In addition, the clinical relevance of NVC findings in DM has not yet been established, as there are inconsistencies among reports, and quantitative analyses have not yet been adequately assessed. Furthermore, red blood cell velocity has not been evaluated, except for our previous report of a small cohort [14] of patients with DM. In this study, we assessed changes in nail-fold capillaroscopy and red blood cell velocity in patients with DM.

Materials and methods

Patients and clinical assessment

Fifty Japanese patients with DM [39 females and 11 males; mean (s.d.) age, 54.7 (14.8) years] who visited Kanazawa University Hospital between 7 October 2007 and 31 July 2010 were included in this study. Forty patients fulfilled the criteria of Bohan and Peter [24, 25], while the remaining 10 did not fulfil the criteria, but fulfilled Sontheimer’s criteria [26], due to the absence of clinical muscle symptoms and presence of subsistent clinical skin eruptions. No patients met the criteria for other rheumatic diseases such as SSc. Mean disease duration was 3.9 (5.1) years. At the time of evaluation, 72% of the patients were already receiving oral predonisolone (PSL) therapy and 26% of the patients were treated with immunosuppressive drugs, including CYC, CSA and tacrolimus, in addition to PSL. A total of 20 healthy subjects [15 females and 5 males; mean (s.d.) age, 51.4 (17.2) years] were also evaluated.

Complete medical histories, physical examinations and laboratory tests were conducted for all patients during the first visit, with limited evaluations during follow-up visits. The patients were diagnosed as having interstitial lung disease (ILD) according to the results of chest radiography, chest CT and pulmonary function testing. Serum Krebs von den Lungen-6 (KL-6) levels as a serum marker of ILD were determined by ELISA as described previously [27]. Presence of internal malignancy was carefully examined using CT, gastrointestinal fibrescope, gallium scintigraphy and other procedures according to need. The local ethics committee (Kanazawa University Hospital, Ishikawa, Japan) approved this study protocol. Informed consent was obtained from each patient.

Immunoprecipitation assays were performed to identify autoantibodies using extracts of the leukaemia cell line K562, as previously described [2]. Using this method, 25.5% of patients were positive for anti-ARS antibody, 18.6% were positive for anti-155/140 antibody and 9.3% were positive for anti-Mi-2 antibody. Anti-SS-A antibody was detected in two patients and anti-NOR 90 antibody, anti-SRP antibody, anti-U1 RNP antibody and anti-Wa antibody were detected in one patient each.

To assess disease activity based on individual organ systems, the Myositis Disease Activity Assessment VAS (MYOACT) portion of the Myositis Disease Activity Tool [28] was used. The MYOACT assessment utilizes separate 100-mm visual analogue scales (VASs) to gauge the physician’s evaluation of disease activity in several discrete domains. Involvement of all non-muscle organ systems (constitutional, cardiac, pulmonary, gastrointestinal, skeletal and cutaneous) was also evaluated using the composite extra-skeletal muscle VAS score. An additional VAS measure, the global VAS score, was used to rate overall disease activity.

NVC pattern classification

We assessed NVC findings using a video capillaroscopy system (CP-1000; Chunichi Denshi, Nagoya, Japan), as previously reported [14]. Classification of NVC patterns was performed according to the criteria of Maricq and colleagues [29, 30]. Diagnostic capillaroscopy patterns were grouped into the following categories, as previously described [31]: (i) normal pattern: homogeneous capillary distribution in the nail-fold plexus without capillary loss (normal medium density: linear 30 capillaries per 5 mm) and no morphological alterations; (ii) scleroderma pattern (defined according to Maricq and colleagues [29, 30], with modifications according to Bergman et al. [32]): two or more of the following abnormalities: enlarged capillaries, haemorrhages (more than two punctate haemorrhages per finger, or confluent haemorrhage areas), disorganization of the normal capillary distribution, moderate or extensive capillary loss (i.e. avascular areas) and tortuous, crossed and/or ramified capillaries; and (iii) non-specific pattern: lack of complete scleroderma pattern criteria. Since it was often difficult to distinguish between the normal and non-specific patterns, these patterns were combined into a normal/non-specific group in this study.

The scleroderma patterns were further classified as previously reported [33], as follows: (i) early NVC pattern: few enlarged/giant capillaries, few capillary haemorrhages, relatively well-preserved capillary distribution and no evident loss of capillaries; (ii) active NVC pattern: frequent giant capillaries, frequent capillary haemorrhages, moderate loss of capillaries with some avascular areas, mild disorganization of the capillary architecture and absent or ramified capillaries; and (iii) late NVC pattern: irregular capillary enlargement, few or absent giant capillaries, absence of haemorrhages, severe loss of capillaries with large avascular areas, severe disorganization of the normal capillary array and frequent ramified/bushy capillaries.

Scoring of NVC findings

A semi-quantitative rating scale to score six capillary parameters (irregularly enlarged capillaries, giant capillaries, haemorrhages, loss of capillaries, disorganization of the vascular array and capillary ramifications) was adopted (0 = no changes, 1 = <33% capillary alterations/reduction, 2 = 33–66% capillary alterations/reduction, 3 = >66% capillary alterations/reduction, per linear mm) according to previous studies [11].

Measurement of red blood cell velocity

We measured red blood cell velocity using the video capillaroscopy system, as previously reported [34]. Mean blood cell velocity was calculated by averaging the results from three capillaries, excluding giant or ramified capillaries, in each ring finger. The examination was performed blindly by the same operator, without knowledge of patients’ clinical conditions or characteristics. Red blood cell speed in nail-fold capillaries was evaluated by frame-by-frame analysis of video data, as described previously [35]: displacement of the edge of the plasma gap was measured on a frame-to-frame basis as it moved in the capillary limb. The velocity was then calculated by relating this displacement to the framing rate (fixed at 30 frames/s) [35].

Statistical analysis

Statistical analyses were performed using JMP® 7.01 Statistical Discovery Software (SAS institute, Cary, NC, USA). Chi-squared test and t-test were used for comparing the frequency and the mean value, respectively. Spearman’s rank correlation coefficient was used to examine the relationship between two continuous variables. Bonferroni correction was performed for multiple comparisons. A P < 0.05 after Bonferroni correction was considered indicative of statistical significance. All data are shown as mean (s.d.), unless otherwise indicated.

Results

NVC pattern classification

Among patients with DM, 74% displayed the NVC scleroderma pattern. Among these patients, the early NVC pattern was observed in 12 (24.0%) patients, the active NVC pattern was noted in 23 (46.0%) patients and the late NVC pattern was recognized in 2 (4.0%) patients. None of the 20 normal control subjects demonstrated a scleroderma pattern.

Association between scleroderma pattern and clinical/laboratory findings

We examined the association between NVC changes and clinical or laboratory features. Patients with the scleroderma pattern had shorter disease duration than patients with the normal NVC pattern, but the difference was not significant (Table 1). The frequencies of muscle weakness, Gottron’s sign and heliotrope rash were higher in patients with the scleroderma pattern than in patients with the normal NVC pattern, but these differences were not significant. In addition, patients with the scleroderma pattern displayed internal malignancies more frequently than patients without the pattern, although this difference was not significant. The presence or duration of RP was not associated with scleroderma pattern. Examination of laboratory findings revealed that patients with the scleroderma pattern displayed elevated serum creatine kinase (CK) levels more frequently than patients without the scleroderma pattern (P < 0.01). Although patients with anti-155/140 antibody had the scleroderma pattern more frequently compared with patients without the antibody, the difference was not statistically significant. The frequency of oral PSL and CSA use was lower in patients with the scleroderma pattern than in patients without it, but the difference was not significant. Other than these results, there were no associations between the NVC scleroderma pattern and clinical or laboratory features in patients with DM. Thus, the scleroderma pattern was most associated with elevated serum CK levels in patients with DM.

Table 1

Association between scleroderma pattern and clinical or laboratory findings

Clinical or laboratory finding	Scleroderma pattern (n = 37)	Normal/ non-specific pattern (n = 13)	P-value
Age, mean (s.d.), years	55.9 (14.5)	51.3 (16.0)	1.0
Sex (male : female)	6 : 31	5 : 8	0.30
Disease duration, mean (s.d.), months	37.0 (62.0)	75.8 (50.7)	0.14
Symptoms, %
Muscle weakness	48.6	30.8	1.0
Gottron's sign	54.1	38.5	0.30
Heliotrope rash	24.3	7.7	0.60
ILDs	43.2	53.8	1.0
Internal malignancy	24.3	7.7	0.22
RP	21.6	15.4	1.0
Laboratory findings, %
Elevated CK	40.5	0	0.0061**
Elevated KL-6	37.8	61.5	0.14
Autoantibodies, %
Anti-ARS antibody	29.7	15.4	0.93
Anti-155/140 antibody	24.3	0	0.15
Anti-Mi-2 antibody	5.4	15.4	0.75
Medications, %
PSL	62.2	100	0.090
CYC	2.7	0	1.0
CSA	8.1	53.8	0.090
Tacrolimus	2.7	0	1.0
MTX	0	7.7	1.0
IVIG	2.7	15.4	1.0

**P < 0.01.

Association between scleroderma pattern and disease activity

To assess disease activity based on individual organ systems, the MYOACT portion of the Myositis Disease Activity Assessment Tool was used. The VAS scales of muscle disease activity were significantly higher in patients with the scleroderma pattern than in patients without it (P < 0.01, Table 2). We examined six aspects of global extra-skeletal muscle disease activity. The total global extra-skeletal muscle disease activity, constitutional disease activity, cutaneous disease activity and skeletal disease activity were higher in patients with the scleroderma pattern than in patients without it, although these differences were not significant. Pulmonary activity was comparable between patients with the scleroderma pattern and patients with the normal/non-specific pattern. Global disease activity, defined as muscle disease activity merged with global extra-skeletal muscle disease activity, was higher in patients with the scleroderma pattern than in patients without it, but the difference was not significant. Thus, the scleroderma pattern was significantly associated with muscle disease activity.

Table 2

Association between scleroderma pattern and myositis disease activity scale

Myositis disease activity scale	Scleroderma pattern (n = 37)	Normal/ non-specific pattern (n = 13)	P-value
Muscle disease activity	20.7 (24.7)	4.6 (7.8)	0.0030*
Global extra-skeletal muscle disease activity	13.7 (7.7)	7.9 (6.7)	0.060
Constitutional disease activity	18.2 (17.0)	6.2 (8.7)	0.29
Cutaneous disease activity	37.0 (23.5)	18.2 (17.8)	0.17
Skeletal disease activity	3.5 (9.2)	0.8 (2.8)	1.0
Gastrointestinal disease activity	2.4 (11.6)	2.3 (6.0)	1.0
Pulmonary disease activity	17.8 (24.2)	20.0 (24.5)	1.0
Cardiac disease activity	3.2 (6.7)	0	1.0
Global disease activity	23.9 (16.6)	11.4 (11.9)	0.054

*P < 0.01. Data are shown as mean (s.d.).

Association between the score of NVC changes and disease activity scales

Among six capillaroscopic parameters, the score of capillary ramifications was excluded in this analysis, since the number of patients with this abnormality was small (12%). The scores for loss of capillaries were significantly associated with the scales of muscle disease activity (r = 0.34; P < 0.05) and global disease activity (r = 0.37; P < 0.05, Table 3). On the other hand, the scores of haemorrhages were significantly associated with the scales of cutaneous disease activity (r = 0.41; P < 0.01, Table 3). However, no parameters were associated with pulmonary disease activity. Thus, there are some specific associations between the scores of NVC changes and muscle, cutaneous and global disease activity scales.

Table 3

Association between the score of NVC change and disease activity scale

Myositis disease activity scale	Irregularly enlarged capillaries, r	Giant capillaries, r	Haemorrhages, r	Loss of capillaries, r	Disorganization of the vascular array, r
Muscle disease activity	0.08	0.11	0.21	0.34*	0.28
Cutaneous disease activity	0.29	0.22	0.41**	0.21	0.30
Pulmonary disease activity	0.03	0.16	−0.15	0.003	0.04
Global disease activity	0.15	0.26	0.19	0.37*	0.33

*P < 0.05, **P < 0.01.

Red blood cell velocity

Mean red blood cell velocity was 0.800 (0.164) mm/s in healthy volunteers (Fig. 1). Patients with DM had a reduced blood velocity of 0.663 (0.204) mm/s, which was 82.9% that of healthy controls; however, this difference was not statistically significant. In addition, the blood velocity values were not significantly different between NVC patterns (early, active, late and normal/non-specific). No significant association between blood velocity and clinical features was found in patients with DM (data not shown). Although there are several patients who showed reduced red blood cell velocity, no specific clinical features were detected in these patients. Thus, red blood cell velocity was not significantly changed in patients with DM.

Fig. 1

Red blood cell velocity in patients with DM and in healthy controls. In patients with DM, the data were also shown dependent on the NVC pattern. Red blood cell velocity was evaluated using frame-to-frame determination of the position of a plasma gap in the capillary. The short bar indicates the mean value in each group.

NVC changes during the longitudinal study

Twelve patients who had a first visit due to active DM were followed up until the disease was stabilized by treatment with immunosuppressive agents, a period of time that averaged 9.2 (8.4) months. The profile of these patients is shown in Table 4. The mean scale of global disease activity was 33.3 (16.1) at their first visit, which was significantly reduced to 11.9 (9.4) after stabilization by treatment (P < 0.001). Of six capillaroscopic parameters, the score of capillary ramifications was excluded from this analysis, since only one patient had this abnormality. The scores of irregularly enlarged capillaries [1.33 (0.89) →0.17 (0.58) mm/s], haemorrhages [1.83 (1.19)→0.17 (0.39) mm/s] and loss of capillaries [0.58 (0.72)→0.08 (0.29) mm/s] were significantly reduced after stabilization of disease (P < 0.01), whereas the scores of giant capillaries [1.92 (0.67)→1.17 (0.39) mm/s] and disorganization of the vascular array [1.17 (1.19)→0.75 (0.87) mm/s] were not significantly changed. Five representative cases are shown in Fig. 2. These pictures demonstrate that irregularly enlarged capillaries, haemorrhages and loss of capillaries are reduced or disappear, and gradually regenerate, after stabilization of disease activity. In contrast, red blood cell velocity was not significantly changed by treatment [0.739 (0.149)→0.653 (0.161) mm/s]. Thus, the NCV abnormalities, especially irregularly enlarged capillaries, and haemorrhages and loss of capillaries likely reflect therapeutic effects in patients with DM.

Fig. 2

Representative NVC images at (A) base line and (B) after treatment, demonstrating how changes can be followed after stabilization of the disease activity in patients with DM. ◂: haemorrhages;  : irregularly enlarged capillaries; #: loss of capillaries.

Table 4

The profile of patients followed until the stabilization of disease activity

Case	Age	Sex	Duration, months	Muscle weakness	ILD	Internal malignancy	RP	Serum CK level, IU/l	Autoantibodies	Initial therapy	Follow-up period, months	Therapy at the second point (inactive)
1	40	Female	2	++	−	−	−	6942	Mi-2	PSL 50 mg, mPSL pulse, CYC, IVIG	15	Rinderon 1 mg, tacrolimus 5 mg
2	42	Male	1	++	−	−	−	4795	Mi-2	PSL 50 mg, mPSL pulse, tacrolimus 3 mg	3	Rinderon 2.5 mg, tacrolimus 3 mg
3	44	Female	3	+	−	+	−	9388		Rinderon 100 mg × 3 days, PSL 30 mg	1	PSL 20 mg
4	51	Female	12	+	+	−	+	345	ARS (Jo-1)	PSL 30 mg	2	PSL 20 mg, tacrolimus 2 mg
5	53	Female	14	−	−	−	−	100	negative	PSL 20 mg	6	PSL 7.5 mg
6	57	Female	3	−	−	−	−	63	U1	PSL 20 mg	5	PSL 10 mg
7	59	Female	9	−	+	−	−	64	Wa	PSL 45 mg, mPSL pulse	3	PSL 25 mg
8	61	Female	30	+	+	+	−	49	ARS (Jo-1)	PSL 50 mg, CSA 750 mg	9	PSL 20 mg, tacrolimus 3 mg
9	61	Male	18	+	+	−	−	8261	ARS (Jo-1)	PSL 50 mg	1.5	PSL 35 mg
10	70	Male	8	+	−	+	−	1700	155/140	PSL 50 mg, mPSL pulse	6	PSL 18 mg
11	73	Female	4	+	+	−	−	126	CADM140	PSL 50 mg, CYC, tacrolimus 2 mg	4	PSL 30 mg, tacrolimus 3 mg
12	82	Female	3	+	−	+	−	169	155/140	PSL 20 mg	2	PSL 15 mg

mPSL: methylprednisolone.

Discussion

Using a video capillaroscopy system, we assessed morphological change and red blood cell velocity in the nail-fold capillaries of DM patients. The NVC scleroderma pattern was frequently detected in patients with DM, and was associated with disease activity, especially muscle disease activity. Among various NVC findings, loss of capillaries was significantly associated with muscle and global disease activities. In addition, haemorrhage was significantly associated with cutaneous disease activity. Findings of irregularly enlarged capillaries, haemorrhage and loss of capillaries were decreased after stabilization of disease activity by treatment. Red blood cell velocity was not significantly reduced in patients with DM compared with normal controls and was not changed by treatment.

Although information regarding nail-fold capillary changes and red cell velocity had been previously available for SSc, such information was not fully established for adult DM. In our study, the NVC scleroderma pattern was found in 74% of DM patients, which was slightly lower than, but comparable to, what we previously reported in SSc patients (84%) [14]. Previous studies have reported significant positive correlation between cutaneous blood flow measured by laser Doppler imaging, and disease severity in adult patients with DM [36]. In contrast, reduced cutaneous blood flow detected by laser Doppler imaging has been reported in patients with PM/DM [19]. However, as far as we know, red blood cell velocity using video capillaroscopy has not been assessed by other groups. We previously reported that patients with SSc showed a 63% decrease in red blood cell velocity compared with healthy controls [34]. In that study, DM patients included as disease controls exhibited slightly but not significantly reduced red blood cell velocities compared with healthy controls. In this study, we confirmed that result in a larger DM population, and assessed the association with clinical features. Although the NVC findings in DM are indistinguishable from those in SSc [12], our findings indicate that the reduction in red blood cell velocity is more modest in DM patients than that in SSc patients. This may reflect somewhat different microcirculation injuries in DM vs SSc.

Our findings indicate that NVC changes are significantly associated with disease activity in patients with DM. Patients with the scleroderma pattern had elevated serum CK levels more frequently and had higher VAS scales of muscle disease activity than patients without the scleroderma pattern. Patients with scleroderma pattern also showed skin symptoms more frequently and elevated cutaneous disease VAS scales compared with patients without scleroderma pattern, although these differences were not significant. On the other hand, the frequency and disease activity of interstitial pneumonia was comparable between patients with the scleroderma pattern and patients without it. Since interstitial pneumonia is often retractable, our findings may at least partly reflect the difficulty of evaluating lung activity. Thus, the current study suggests that NVC change is associated with disease activity, especially muscle disease activity.

Our study identified several disparities between DM patients who displayed the scleroderma pattern and those who did not. For example, scleroderma pattern DM patients had shorter disease duration than DM patients without the scleroderma pattern, although the difference was not significant. Furthermore, patients without scleroderma pattern were receiving PSL and CSA more frequently than patients with scleroderma pattern, although these difference were not significant. These findings likely reflect the fact that patients with short disease duration tend to have active disease, whereas most patients with long disease duration are stable with treatment. In fact, the clinical features at their active phase (before treatment) were not significantly different between patients treated with PSL or CSA and patients not receiving treatment (data not shown). In addition, DM patients with the scleroderma pattern had internal malignancies more frequently than DM patients without the scleroderma pattern, although the difference was not significant. Consistent with this, anti-155/140 autoantibody, which is commonly detected in DM patients with internal malignancy, tended to be more frequently detected in DM patients with the scleroderma pattern than in patients without it. Since DM patients with either anti-155/140 antibody or internal malignancy typically exhibit cutaneous eruption and myositis without lung involvement [3, 4], such associations may be due to the cutaneous and muscle disease activity in these patients.

Importantly, NVC changes were improved by disease stabilization in DM patients during the follow-up period. Among NVC changes, irregularly enlarged capillaries, haemorrhages and loss of capillaries were significantly reduced after stabilization of disease activity (Fig. 2). Therefore, monitoring these changes will likely be useful in evaluating disease activity and therapeutic efficacy. On the other hand, it has been reported that capillary loss is associated with progression of SSc and generally of the microvascular damage in secondary Raynaud’s syndrome, at least in SSc [37–39]. In SSc patients, giant capillaries and haemorrhages were not considered critically important in the evaluation of SSc microangiopathy, as these abnormalities are evident only in the early stages of the disease, and then disappear or become rare in the advanced stages [33, 40]. Thus, our study demonstrates that the significance of each NVC change is different in some degree between DM and SSc.

A previous study of adult-onset PM and DM patients found that RP, arthritis and pulmonary involvement were associated with increased numbers of enlarged capillary loops and more severe avascular lesions [18]. In that study, the severity of the observed abnormalities did not correlate with the occurrence of malignancy or active myositis, but tended to decline with prolonged disease remission [18]. In a recent study including 53 adult patients with inflammatory myopathy, disease activity and severity were both significantly associated with alterations in capillary morphology [20]. Furthermore, marked abnormalities of capillaries were significantly associated with the involvement of internal malignancy or ILD. There are also some reports regarding NVC findings in JDM. One study found that NVC abnormalities are associated with skin involvement in patients with JDM [22]. Another study demonstrated that capillary loss was associated with skin involvement in JDM [21]. A prospective study involving 13 JDM patients demonstrated that capillary dropout was most frequently correlated with disease activity [23]. Longer duration of untreated disease and severe skin lesions were associated with capillary reduction in JDM [41]. Regarding associations with autoantibody, anti-Jo-1 antibody was associated with reduced capillary density [17]. Thus, our findings show some discrepancies with previous findings in patients with inflammatory myopathy. The main cause is likely due to the heterogeneity of inflammatory myopathy, and this makes the case for studying adult DM exclusively. Most previous papers have assessed inflammatory myopathy, both DM and PM, whereas our study is restricted to DM. For example, it has been reported that microhaemorrhages and capillary enlargement appear to represent the characteristic NVC pattern in patients with DM, but not in those with PM [20]. Ethnic differences also affect the results. Nonetheless, our results in DM, along with previous findings in inflammatory myopathy, are at least consistent in finding that general disease activity and severity are associated with prominent morphological changes.

Thus, our findings, together with previous reports, indicate that NVC findings are useful for diagnostic purposes, as well as for assessment of disease activity and response to treatment in patients with DM. Nonetheless, this study has some limitations. Although DM is rare, the number of patients analysed was not large and the study was restricted to Japanese individuals. Furthermore, most patients were already stable due to treatment at the time of evaluation. Therefore, further prospective multicentre studies using larger patient populations will be needed to confirm our results.

Acknowledgements

We thank Tomoko Hayashi for her technical assistance.

Funding: This work was supported by funds for research on intractable diseases from the Ministry of Health, Labour and Welfare of Japan.

Disclosure statement: The authors have declared no conflicts of interest.