Use of nailfold capillaroscopy for evaluation of disease activity in juvenile dermatomyositis: Results of a two-center retrospective study

Abstract

BACKGROUND

Juvenile dermatomyositis (JDM) is a rare, immune-mediated children disease that affects both skeletal muscles and skin. Inflammatory vasculopathy is a key part of the JDM pathogenesis. Nailfold capillaroscopy (NFC) is a non-invasive method for microvasculature assessment. Despite the pathogenic relevance of microangiopathy in JDM, the diagnostic utility of NFC in JDM remains insufficiently defined.

AIM

To quantitatively assess capillaroscopic dynamics and their association with clinical and laboratory activity markers in JDM.

METHODS

This retrospective double-center cohort study included 52 children with confirmed JDM, who were followed at the clinics of Sechenov University and Saint Petersburg State Pediatric Medical University. Patients were divided into active and inactive JDM groups based on the modified disease activity score. Demographic, clinical, laboratory, and instrumental data were analyzed. NFC evaluation was performed using standardized criteria developed initially for adult patients with systemic sclerosis and Raynaud’s phenomenon. Statistical analysis was conducted using Windows v.13 (StatSoft).

RESULTS

Active and inactive JDM groups were demographically comparable (predominantly female, median onset age 6 years). NFC revealed marked differences: Active patients showed higher rates of reduced capillary density (96.2% vs 60.9%, P = 0.003), giant capillaries (73.1% vs 13.0%, P < 0.001), perivascular edema (73.1% vs 17.4%, P < 0.001), and greater maximal apical loop width. Disease activity positively correlated with giant capillaries (r = 0.65), perivascular edema (r = 0.60), and reduced capillary density (r = 0.46). These findings confirm that specific NFC features are strongly associated with clinical disease activity, as measured by the disease activity score.

CONCLUSION

Children with active JDM demonstrated more frequent NFC changes, indicating that NFC may play a potential role in monitoring disease activity. Further studies are needed to determine the diagnostic accuracy and prognostic relevance of this approach.

Key Words: Nailfold capillaroscopy; Juvenile dermatomyositis; Inflammatory myopathy; Disease activity; Children

Core Tip: This study demonstrates that specific nailfold capillaroscopy (NFC) abnormalities - notably giant capillaries, perivascular edema, and severely reduced capillary density - are strongly correlated with clinical disease activity in juvenile dermatomyositis. We quantitatively prove that these NFC features are significantly more prevalent in active juvenile dermatomyositis, positioning NFC as a valuable, non-invasive tool for monitoring microvascular damage and assessing disease activity in children, extending its traditional use to scleroderma spectrum disorders.

INTRODUCTION

Juvenile dermatomyositis (JDM) is a rare, systemic, immune-mediated children disease characterized by progressive involvement of striated muscles and specific cutaneous manifestations[1]. The prevalence of JDM ranges from 2.5 cases to 4.0 cases per million children, underscoring its status as an orphan disease[2,3]. Despite its low prevalence, the disease usually has a severe course and, without adequate therapy, leads to a significant disability, a decrease in quality of life, and a high risk of mortality.

Accurate assessment of disease activity plays a crucial role in the management of JDM, as it determines prognosis and guides therapeutic strategy[4-6]. Several tools are used in clinical practice for this purpose, including combined indices [such as disease activity score (DAS), juvenile dermatomyositis activity index] that assess both muscular and cutaneous components of the disease, as well as specialized scales for skin lesions (skin DAS, cutaneus assessment tool , cutaneous dermatomyositis disease area and severity index), which are particularly important since the severity of skin manifestations often correlates with unfavorable outcomes[7-10].

For last years, methods aimed at assessing the deep pathogenic mechanisms of JDM have been introduced actively in clinical practice. For instance, biomarkers such as type I interferon-α levels demonstrate close correlation with disease activity both at onset and during long-term follow-up[11,12], and muscle ultrasound is highly promising for non-invasive monitoring of inflammatory processes[13-17].

Nailfold capillaroscopy (NFC) is a non-invasive method for visual assessment of microcirculation, which can be used in various pathological conditions[18-21]. Given that the pathogenesis of JDM is based on a “humoral attack” on the microcirculation, NFC allows direct evaluation of a key pathological process. Characteristic capillaroscopic patterns, including decreased capillary density, giant capillaries, microhemorrhages, and disorganization of the vascular network architecture, are observed in most JDM patients and are associated with indicators of inflammatory activity[22-24].

Despite the widely recognized pathogenetic significance of microangiopathy in JDM, the potential of quantitative assessment of capillaroscopic changes for monitoring disease activity and predicting treatment response remains insufficiently studied. Most studies are descriptive, and there is a lack of unified protocols for integrating NFC data into clinical scales.

This study aimed at quantitatively assessment of the dynamics of capillaroscopic parameters in JDM patients undergoing standard therapy and analysis of their correlation with traditional clinical and laboratory markers of disease activity.

MATERIALS AND METHODS

Study design

In a double-center retrospective cohort study, 49 children with confirmed JDM undergoing NFC were included, observed across clinics of Sechenov University and Saint Petersburg State Pediatric Medical University between October 2024 and May 2025.

Inclusion criteria: (1) Age under 18 years; (2) Confirmed diagnosis of JDM, according to the 2017 European League Against Rheumatism/American College of Rheumatology classification criteria for juvenile idiopathic inflammatory myopathies[25]; and (3) Complete clinical and capillaroscopy data available.

Exclusion criteria: (1) Incomplete medical records preclude reliable disease activity assessment; and (2) Presence of overlapping connective tissue diseases, or secondary Raynaud’s phenomenon.

Assessments

Demographic (age at the moment of NFC performance, sex, age at disease onset, disease duration), clinical (systemic inflammatory symptoms, skin and muscle involvement, calcinosis, extramuscular organ involvement), laboratory: C-reactive protein, erythrocyte sedimentation rate, hemoglobin, erythrocytes, leukocytes, platelets, alanine aminotransferase, aspartate aminotransferase, creatine phosphokinase, lactate dehydrogenase (LDH), antinuclear antibodies (ANA), and NFC data were taken from medical records.

Patients were divided into two groups based on disease activity at the time of NFC: Patients with active JDM (n = 26) and patients with inactive JDM (n = 23). Given the retrospective nature of the study, disease activity was assessed using a modified version of the DAS (DASmod), which was adapted from previously validated tools[26]. The DASmod incorporates clinical parameters, such as skin involvement, muscle weakness, and physician global assessment. Modifications were made to accommodate retrospective data collection, including the substitution of missing visual analogue scale scores with documented clinical severity grades where available. This modified score has demonstrated good correlation with the original DAS in prior validation studies[26]. Active disease was defined as a DASmod score of 3 or higher, a threshold associated with clinically meaningful disease activity in pediatric populations. The detailed DASmod assessment is depicted in Table 1.

Table 1 Modified disease activity score index.

Parameters	Scored item	Options	Score
Muscle DAS (0-7)	Muscle strength	No muscle weakness (gMMT 68-70, MMT8 78-80)	0
		Minimal muscle weakness (gMMT 63-67, MMT8 72-77)	1
		Moderate muscle weakness (gMMT 56-62, MMT8 64-71)	2
		Severe muscle weakness (gMMT 0-55, MMT8 0-63)	3
	Functional status	No limitation on activity	0
		Activities limited at the extracurricular level	1
		Activities limited at the school/work level	2
		Activities limited at the self-care level	3
	Arthritis	Absent	0
		Present	1
Skin DAS (0-5)	Erythema	No erythema	0
		Local erythema (joints and face)	1
		Widespread erythema	2
	Heliotropic rash	Absent	0
		Present	1
	Gottron papules	Absent	0
		Present	1
	Vasculitis	Absent	0
		Present (nailfold abnormality)	1
Total score (0-12)

DAS: Disease activity score; gMMT: Global manual muscle test; MMT8: Manual muscle test 8.

NFC

NFC was performed on the volar surface of the distal phalanges of the second to fifth fingers of both hands using a video capillaroscope with magnification ≥ 200 ×. Images were acquired and evaluated by trained rheumatologists who were blinded to the clinical data. Capillaroscopy findings were assessed using standardized criteria established for systemic sclerosis and Raynaud's phenomenon[27]: Normal capillary density was defined as ≥ 7 capillaries per millimeter; capillaries with an apex diameter > 20 μm were classified as dilated, and those > 50 μm as giant capillaries. Additional features evaluated included microhemorrhages, avascular areas, capillary architectural disorganization, perivascular edema, and extravasations.

Statistical analysis

Statistical analysis was performed using StatSoft’s Windows v.13. Normality of distribution was tested using the Shapiro-Wilk test. Normally distributed continuous variables are presented as mean ± SD, with 95% confidence intervals (CIs). Non-normally distributed variables are reported as median and interquartile range, median (Q1-Q3). Categorical variables are expressed as absolute numbers and percentages, with 95%CIs calculated using the Clopper-Pearson method.

Group comparisons were performed using Student’s t-test for normally distributed data with equal variances, or the Mann-Whitney U test for non-normally distributed data. For categorical variables, χ² Pearson’s test was used for 2 × 2 and larger contingency tables when expected frequencies were ≥ 10; Fisher’s exact test was applied when expected frequencies were < 10. The effect level for binary outcomes was expressed as an odds ratio (OR) with 95%CI. The Haldane-Anscombe correction was applied in cases with zero cell counts. A two-sided P value < 0.05 was considered statistically significant. Spearman’s rank correlation coefficient was used to assess the relationship between variables.

RESULTS

Clinical and laboratory characteristics of the groups

Demographic and baseline clinical characteristics were comparable between the two groups (Table 2). Female patients predominated in both groups (65.4% in Active JDM group 1 vs 56.5% in Inactive JDM group). The median age at disease onset was 6 [interquartile range (IQR): 5-8] years, and the median time from symptom onset to diagnosis was 3 (IQR: 2-6) months. The median disease duration at the time of NFC was 40 (IQR: 13-67) months and did not differ significantly between the groups (P = 0.161). The mean DASmod at diagnosis did not differ significantly between groups [8 (IQR: 6-9), P = 0.21], indicating similar initial disease severity.

Table 2 Clinical and laboratory characteristics of the studied groups (data from the day of capillaroscopy), n (%).

Parameters	Active JDM (n = 26)	Inactive JDM (n = 23)	P value
Age at onset, years, median (IQR)	6.0 (5.0, 8.0)	6.0 (4.5, 9.0)	0.755
Age at capillaroscopy, years, mean ± SD (95%CI)	9.69 ± 4.34 (7.94-11.45)	11.4 ± 3.9 (9.7-13.1)	0.156
Disease duration before diagnosis, months, median (IQR)	3.5 (2.0-5.8)	3.0 (2.0-5.5)	0.976
Disease duration before capillaroscopy, months, median (IQR)	36.0 (11.5-52.5)	56.0 (15.0-86.0)	0.161
DAS mod at diagnosis, mean ± SD (95%CI)	7.8 ± 1.9 (7.0-8.6)	7.7 ± 1.8 (6.9-8.4)	0.772
DAS mod at capillaroscopy, median (IQR)	5.0 (4.0, 7.0)	1.0 (1.0, 2.0)	< 0.001
Clinical features
Malaise	13 (50.0)	2 (8.7)	0.002
Headache	2 (7.7)	0 (0.0)	0.491
Fever	3 (11.5)	0 (0.0)	0.237
Erythema	21 (80.8)	8 (34.8)	0.001
Heliotrope rash	20 (76.9)	3 (13.0)	< 0.001
Gottron papules	25 (96.2)	7 (30.4)	< 0.001
Raynaud’s syndrome	8 (30.8)	2 (8.7)	0.080
Muscle weakness	17 (65.4)	5 (21.7)	0.002
Activities limited	11 (42.3)	3 (13.0)	0.030
Arthritis	3 (11.5)	4 (17.4)	0.612
Arthralgia	3 (11.5)	4 (17.4)	0.692
Calcinosis	3 (11.5)	4 (17.4)	0.692
Cardiovascular system disorders	3 (11.5)	1 (4.3)	1.000
Digestive system disorders,	1 (3.8)	0 (0.0)	1.000
Laboratory features
HGB g/L, mean ± SD	130.9 (13.8)	130.4 (12.7)	0.905
RBC (1012/L), mean ± SD	4.7 (0.4)	4.7 (0.5)	0.847
WBC (109/L), median (IQR)	6.2 (5.4, 8.8)	6.5 (5.8, 8.0)	0.667
PLT (109/L), mean ± SD	339.6 (67.7)	350.8 (72.7)	0.579
ESR (mm/hour), median (IQR)	8.5 (4.3, 17.0)	8.0 (5.0, 9.5)	0.393
Ferritine (ng/mL), median (IQR)	39.7 (32.0, 72.0)	40.0 (33.4, 58.1)	0.942
CRP (mg/L), median (IQR)	1.2 (0.2, 2.7)	1.4 (0.5, 1.8)	0.748
ALT (IU/L), median (IQR)	27.5 (13.5, 47.0)	16.0 (11.5, 21.5)	0.071
AST (IU/L), median (IQR)	34.5 (22.3, 43.8)	26.0 (23.5, 32.0)	0.061
Common protein (g/L), median (IQR)	70.0 (67.3, 73.0)	69.0 (66.5, 71.0)	0.622
Albumin (g/L), mean ± SD	40.6 (4.0)	43.3 (2.9)	0.036
Bilirubin (µmol/L), median (IQR)	7.0 (4.8, 10.3)	8.8 (5.5, 11.7)	0.312
CPK (IU/L), median (IQR)	107.0 (79.0, 198.0)	77.0 (54.5, 89.5)	0.022
LDH (IU/L), median (IQR)	260.0 (224.0, 341.0)	220.0 (198.5, 233.5)	0.006
Fibrinogen (g/L), median (IQR)	2.6 (2.4, 2.7)	2.8 (2.6, 3.2)	0.119
D-dimer (ng/mL), median (IQR)	254.5 (144.3, 421.8)	228.0 (174.0, 239.0)	0.903
RF (IU/mL), median (IQR)	4.6 (3.3, 6.3)	6.0 (4.8, 6.3)	0.284
Treatment at capillaroscopy
GC	22 (84.6)	12 (52.2)	0.028
Pulse-GC	8 (30.8)	0 (0.0)	0.005
GC dose at capillaroscopy, mg/kg/day in prednisolone eq, median (IQR)	0.4 (0.1, 1.0)	0.2 (0.0, 0.3)	0.019
MTX	16 (61.5)	12 (52.2)	0.623
MMF	2 (7.7)	1 (4.3)	0.623
IVIG	5 (19.2)	0 (0.0)	0.054

JDM: Juvenile dermatomyositis; IQR: Interquartile range; CI: Confidence interval; DAS: Disease activity score; HGB: Haemoglobin; RBC: Red blood cells; WBC: White blood cells; PLT: Platelets; ESR: Erythrocyte sedimentation rate; CRP: C-reactive protein; ALT: Alanine aminotransferase; ASTA; Aspartate aminotransferase; CPKA; Creatine phosphokinase; LDH: Lactate dehydrogenase; RFA; Rheumatoid factor; GC: Glucocorticoids; MTX: Methotrexate; MMF: Mycophenolate mofetil; IVIG: Intravenous immunoglobulin.

As expected, patients with active JDM more frequently showed clinical signs of disease activity at the time of NFC, including malaise (P = 0.002), erythema (P = 0.001), heliotrope rash (P < 0.001), Gottron’s papules (P < 0.001), muscle weakness (P = 0.002), and limitation of activity (P = 0.030). In contrast, no significant difference was observed between groups in the prevalence of arthritis, arthralgia, calcinosis, or involvement of the cardiovascular or gastrointestinal systems.

Among laboratory parameters, no significant differences were found in complete blood count indices (hemoglobin, erythrocytes, leukocytes, and platelets) or systemic inflammatory markers (erythrocyte sedimentation rate and C-reactive protein) between the two groups. However, as anticipated, patients with active JDM had significantly higher levels of creatine kinase and LDH compared to those with inactive disease (P < 0.001 for both). Notably, median values of creatine kinase and LDH remained within the normal reference range in both groups, likely reflecting partial treatment response or subclinical muscle involvement.

Interestingly, ANA positivity was significantly prevalent in children with active JDM at the time of NFC (53.8% vs 34.7%) with a median level of 480.0 (120.0, 1600.0) vs 160.0 (0.0, 320.0) (P = 0.046), as well as at disease onset (P = 0.019), suggesting a potential association between ANA levels and disease activity.

Regarding treatment at the time of NFC, patients with active JDM more frequently received systemic glucocorticoids (GCs), including pulse intravenous methylprednisolone (P = 0.005). The median daily oral GC dose was significantly higher in group 1 (0.40 mg/kg/day) vs group 2 (0.20 mg/kg/day) (P = 0.019). Intravenous immunoglobulin was administered during the NFC period in 5 patients with active JDM. These differences in therapy are consistent with higher disease activity, necessitating intensified immunosuppressive treatment.

Capillaroscopic findings according to JDM activity

Significant differences in NFC patterns were observed between patients with active and inactive JDM (Table 3).

Table 3 Capillaroscopic features of the studied groups, n (%).

Parameters	Active JDM (n = 26)	Inactive JDM (n = 23)	P value
Age at capillaroscopy, mean ± SD	9.7 (4.3)	11.4 (3.9)	0.156
Duration of disease before capillaroscopy, months, median (IQR)	36.0 (11.5, 52.5)	56.0 (15.0, 86.0)	0.161
Density average, n/mm, mean ± SD	3.6 (1.5)	6.3 (1.3)	< 0.001
Low density	25 (96.2)	14 (60.9)	0.003
Maximal arterial dimension, μm, median (IQR)	43.0 (29.6, 54.5)	21.6 (18.2, 25.7)	< 0.001
Maximal apical dimension, μm, median (IQR)	65.1 (46.6, 96.0)	31.9 (24.9, 40.3)	< 0.001
Maximal venous dimension, μm, median (IQR)	56.4 (39.6, 75.5)	26.3 (21.8, 37.0)	< 0.001
Dilated capillaries	26 (100.0)	22 (95.7)	0.469
Giant capillaries	19 (73.1)	3 (13.0)	< 0.001
Abnormal morphology	25 (96.2)	19 (82.6)	0.173
Microhaemorrhages	17 (65.4)	10 (43.5)	0.124
Perivascular extravasates	10 (38.5)	3 (13.0)	0.057
Perivasular edema	19 (73.1)	4 (17.4)	< 0.001

JDM: Juvenile dermatomyositis; IQR: Interquartile range.

A marked reduction in nailfold capillary density (NCD) was observed in 96.2% of patients with active JDM, compared to 60.9% in the inactive group (P < 0.001). The mean NCD was 3.6 capillaries/mm in group 1 capillaries/mm vs 6.3 capillaries/mm in the Inactive JDM group (P < 0.001), Figure 1A.

Figure 1 Average density and maximal apical dimension analysis by group. A: Analysis of average density by the group; B: Maximal apical dimension analysis by the group. JDM: Juvenile dermatomyositis.

Morphological abnormalities were more pronounced in active JDM. Giant capillaries were detected in 73.1% of patients with active JDM, compared to 13.0% in patients with inactive JDM (P < 0.001). Perivascular edema was also significantly more common in the active group (73.1% vs 17.4%, P < 0.001). Microhemorrhages were present in both groups, but there was no significant difference between them.

The maximal apical loop width was significantly greater in patients with active JDM: Median 65.10 μm (IQR: 46.58-96.04) vs 31.90 μm (IQR: 24.85-40.25) in the inactive group (P < 0.001) (Figure 1B).

Active JDM was characterized by a predominance of capillaries with abnormal morphology, most commonly bushy capillaries, observed in 57.1% of children (a feature typical of the late scleroderma pattern). In contrast, such changes were present in only 34.6% of patients in Group 2 (Figure 2).

Figure 2 Distribution of patients (%) according to the predominant capillary morphological changes. JDM: Juvenile dermatomyositis.

Spearman’s correlation analysis revealed significant positive associations between disease activity (DASmod) and several capillaroscopic features: Presence of giant capillaries (r = 0.65, P < 0.05), perivascular edema (r = 0.60, P < 0.05), and reduced capillary density (r = 0.46, P < 0.05). These findings suggest that specific NFC abnormalities correlate with clinical disease activity in JDM.

Correlations between certain capillaroscopic and clinical features were also evaluated. There was no association between capillaroscopic and systemic features of the disease. However, some strong correlations were found between NFC findings and skin features of the disease. They are listed in Table 4.

Table 4 Correlations between capillaroscopic and skin features.

Parameters	Low density	Dilated capillaries	Giant capillaries	Abnormal morphology	Microhaemorrhages	Perivascular extravasates	Perivasular edema
Erythema	0.40a	0.17	0.25	0.13	0.25	0.22	0.37a
Heliotrope rash	0.37a	0.14	0.47	0.05	0.19	-0.01	0.34a
Gottron papules	0.48a	0.20	0.49	0.18	0.20	0.15	0.51a
Raynaud’s syndrome	0.13	-0.29a	-0.05	0.17	0.15	0.04	0.23

^aP < 0.05.

DISCUSSION

Our findings suggest that NFC may serve as a promising tool for evaluating disease activity in JDM. These results demonstrate clear association between specific microvascular abnormalities and clinical disease activity, supporting the potential role of NFC in monitoring JDM. They also corroborate previous studies showing that reduced capillary density (NCD) is associated with disease activity[23,28,29]. In our cohort, decreased NCD was observed in all patients with JDM, irrespective of disease activity; however, this reduction was significantly more pronounced in those with active disease. In addition to reduced NCD, our study identified significant association between disease activity and the presence of giant capillaries and perivascular edema. These microangiopathic changes may serve as valuable biomarkers for both diagnosis and longitudinal monitoring of disease progression, offering non-invasive insights into microcirculatory health through NFC.

Of particular importance is the high prevalence of NCD (60.9%) and other abnormalities in patients with clinically inactive JDM, which demonstrates that clinical remission does not equate to the complete recovery of the microvasculature. The persistence of these changes represents irreversible damage to the endothelium and vessel architecture. These data compellingly confirm that NFC identifies a JDM microvasculopathy that may endure longer than the manifest clinical signs.

A higher prevalence of neovascular “bushy” capillaries was observed in patients with active JDM, consistent with the study by Barth et al[30], which demonstrated a statistically significant association between this capillaroscopic pattern and disease activity in 58 children with JDM.

Some studies suggested correlation between specific myositis-associated autoantibodies and the extent of capillary density loss[23,31]. Although our study did not assess antibody profiles, this potential immunogenetic link warrants further investigation in larger, serologically stratified cohorts.

Conflicting evidence exists in the association between NFC abnormalities and systemic complications, such as calcinosis or interstitial lung disease[30,32]. In our cohort, no such correlations were observed, which may reflect differences in patient selection, disease duration, or treatment strategies, among others. Further multicenter studies with standardized outcome assessments are needed to clarify these relationships.

The impact of disease duration - and particularly diagnostic delay - on microvascular changes was previously debated[32,33]. In our study, we analyzed NFC findings based on time from symptom onset to diagnosis (≤ 6 months: 36 patients; > 6 months: 13 patients) and total disease duration at assessment (≤ 12 months: 11 patients; > 12 months: 38 patients). No significant differences in NFC abnormalities were found between these groups, suggesting that NFC changes may manifest early and persist regardless of disease duration. However, we did observe a significantly higher prevalence of arthritis in children with a diagnostic delay of more than 6 months (69.2% vs 30.6%, P = 0.022). The odds of developing arthritis were 5.114 times higher in the late-diagnosis group, with a statistically significant OR (OR = 0.196; 95%CI: 0.049-0.773). This finding underscores the clinical importance of early diagnosis and intervention, not only to mitigate skin and muscular manifestations but also to reduce the risk of articular complications.

Furthermore, we found no significant correlation between NFC abnormalities and specific medications or treatment regimens. While the potential confounding effect of higher GC use in the active disease group must be acknowledged, its independent impact on the specific NFC pattern of JDM is likely limited. It is indirectly supported by the fact that other immune-mediated conditions (e.g., systemic lupus erythematosus), which are also treated with high-dose GCs, do not typically exhibit the pronounced pattern of capillary dropout and giant capillaries characteristic of JDM[34]. Observed microvascular changes are intrinsically linked to JDM pathogenesis rather than being a mere treatment artifact; however, the potential effect of cumulative drug exposure on long-term capillary recovery warrants further investigation.

In a prospective follow-up study by Nascif et al[35] on children with JDM and overlap syndromes, it was demonstrated that NFC patterns evolve with disease activity: 12 out of 13 assessments during active disease identified a scleroderma-like pattern, compared to 8 out of 13 showing normal capillaroscopic findings during remission (P = 0.01). Although our study was not designed for systematic longitudinal monitoring, serial NFC examinations in a subset of patients revealed clear improvements in microvascular parameters following the initiation of immunosuppressive therapy - specifically, increased capillary density and reduced numbers of giant capillaries and microhemorrhages (Figure 3). These findings suggest that NFC may serve as a sensitive, non-invasive tool for monitoring treatment response and disease activity in JDM. However, these observations require validation in larger, prospective, longitudinal studies.

Figure 3 Dynamics of microvascular changes in a 15-year-old girl with juvenile dermatomyositis before and after therapy. The black line at the top of each image represents 1 mm. Note the increase in capillary density from 2-3/mm to 4-5/mm, the disappearance of microhemorrhages and giant capillaries, and the emergence of a bushy capillary pattern, indicating neovascularization. A: Left hand, second finger before therapy; B: Left hand, second finger after four months of Janus kinase inhibitor therapy; C: Right hand, second finger before therapy; D: Right hand, second finger after 4 months of Janus kinase inhibitor therapy.

Limitations

Our study has some limitations. They are due to the retrospective type, double-center design, relatively small sample size, partially missing data, and different study time intervals. The age of patients varied, with a wide range of periods between disease onset and capillaroscopy. Also, NFC was performed using different capillaroscopes. We considered a relatively limited list of laboratory studies. Almost no links were found between laboratory parameters and capillaroscopy results. Using other laboratory methods, these links could have been found.

CONCLUSION

Reduced NCD, the presence of giant capillaries, and perivascular edema were significantly prevalent in children with active JDM. These findings indicate that NFC may serve as a valuable non-invasive tool for assessing disease activity. Further research is warranted to evaluate the diagnostic accuracy and prognostic utility of NFC in the management and monitoring of JDM.

ACKNOWLEDGEMENTS

The authors sincerely thank Professor Natalia A. Geppe, Head of the Department of Children’s Diseases at Sechenovskiy University, for her invaluable support and facilitation of this study.