Lower NK cell numbers in children with untreated juvenile dermatomyositis during the COVID-19 pandemic

Abstract

RNA viruses have been posited as triggers for Juvenile Dermatomyositis (JDM). The COVID-19 pandemic proved a unique opportunity to observe the effect of a novel RNA virus on JDM incidence and phenotype. We found the incidence of JDM increased from average of 6.9 cases per year from 2012 to 2019 to 9 cases per year from 2020 to 2021. We compared markers of disease activity in the patients diagnosed with JDM prior to and during the pandemic and found that patients diagnosed with JDM during the pandemic had significantly lower average NK cell counts 90.75(± 76) vs 163(±120) (P = 0.038) and NK cell percentage 3.63% (±2.3) vs. 6.6% (±4.1), (P = 0.008). Other markers of JDM did not significantly change. This study suggests that COVID-19 may be a viral trigger for JDM in selected cases and that NK cell dysregulation may be of particular interest in future research of virally triggered JDM.

1. Introduction

Juvenile Dermatomyositis (JDM) is a rare systemic pediatric autoimmune disease characterized by inflammation targeting skin and muscle. The exact cause of JDM remains unknown; however, a combination of genetic susceptibility and environmental triggers is the prevailing theory [1]. Prior research has shown that genetic changes including TNFa308 polymorphisms are associated with JDM severity [2]. HLA studies have shown that the presence of HLA DRB1∗0301 increases the risk for JDM [3]. Viral triggers for JDM have been reported. One study found the more than half of patients with newly diagnosed JDM had upper respiratory symptoms prior to their diagnosis [4]. Another study showed that DS RNA exposure to epithelial cells generated similar gene expression as found in JDM muscle biopsies implicating RNA viruses as possible triggers for JDM [5].

COVID-19 is a single strand RNA virus and we hypothesized that it may trigger JDM in a susceptible host. We observed two cases of JDM after COVID-19 infection. A 5-year-old previously healthy female developed upper respiratory infection (URI) symptoms and was diagnosed with COVID-19 3 months prior to rheumatology presentation. She had Gottron's sign, heliotrope rash, and weakness resulting in admission where she had a decreased Childhood Myositis Assessment Scale (CMAS) of 24 accompanied by an elevated CK, AST, ALT, LDH, and aldolase. A T-2 weighted image on MRI showed diffuse myositis. Myositis specific antibody (MSA) testing revealed a positive MJ antibody. She was diagnosed with JDM and started on intravenous steroids, methotrexate, hydroxychloroquine, and IVIG with improvement.

A second case of JDM was diagnosed in a previously healthy 4-year-old female at 3 months after a COVID-19 diagnosis. She also presented with weakness, Gottron's sign, and elevated muscle enzymes. She had an MRI showing myositis. Her MSA showed a positive P155/140 antibody, and her symptoms improved with hydroxychloroquine, IV steroids, IVIG and methotrexate.

Although both patients presented somewhat typically, COVID-19 has been known to cause unique autoimmune disease. There are small studies and case reports showing that COVID may serve as a trigger for JDM [6,7]. COVID has been implicated in development of the unique autoimmune disease multisystem inflammatory system of childhood (MIS-C) [8]. Here we hypothesize that COVID triggers JDM and may lead to unique presentations or changes in the JDM phenotype. Determining whether COVID was the exact trigger for JDM proved difficult due to patients with asymptomatic infections, the lack of testing early in the pandemic, long lead times from trigger to disease, and IVIG clouding the use of antibody testing. Despite this determining whether the rate or characteristics of JDM changed during the COVID pandemic remains clinically useful. Furthermore, an increase in JDM rate during the pandemic would indirectly further bolster the posited association of RNA viruses as a trigger for JDM. Here we performed a retrospective observation study aimed to determine whether there was increase in JDM rate during the COVID-19 pandemic or if there were any deviations from our usual clinical and laboratory data in patients diagnosed with JDM during the pandemic compared to patients diagnosed with JDM prior to the pandemic.

2. Materials and methods

2.1. Subjects

This retrospective chart review study was approved by the Institutional Review Board (IRB) at Ann & Robert H. Lurie Children's Hospital of Chicago (IRB 2012–14,858) on October 4, 2021 and was reviewed annually with the last renewal on January 16th 2022. All patients were recruited from the Cure JM biorepository and registry (IRB 2010–14,117), which houses clinical data, laboratory data, and patient samples obtained throughout the disease course of JDM.

All children who were seen at Ann & Robert H. Lurie Children's Hospital between 2010 and 2021, who met the Bohan and Peter criteria for definite or probable JDM, and signed informed consent (in accordance with the Helsinki agreement), were included in the study. Patients with a diagnosis of overlap syndrome or overlap myositis were excluded from the study, as were DM patients over the age of 18 at diagnosis.

We identified 18 patients with a new diagnosis of JDM between 2020 and 2021 who were designated as the post-pandemic group. Recruitment into the registry was hampered by patient sequestration secondary to the pandemic; ultimately 10 of these patients were enrolled in the Cure JM Registry and included in the analysis. This population was compared to a total of 69 patients diagnosed with JDM between January 1st 2010 and December 31st 2019 who were included in the registry and were designated as the pre- pandemic group. The pre pandemic and post pandemic group were similar with full demographics as below in Table 1 . Race and ethnicity were determined by questionnaire with a fixed set of categories.

Table 1.

Demographics and MSA Distribution of 79 JDM Patients.

	Post-Pandemic	Pre-Pandemic	p-Value
Number of subjects	10	69
Age at onset at First Visit	7.47 (±4.23)	7.50(±4.52)	0.985
Sex
Female	9 (90%)	52 (75%)	0.442
Male	1 (10%)	17 (25%)
Race/ethnicity
White, Non-Hispanic	6 (60%)	48 (74%)	0.54
White, Hispanic	3 (30%)	10 (15%)
African American	0 (0%)	2 (3%)
Asian	0 (0%)	4 (5%)
Other/Unspecified	1 (10%)	1 (3%)
Myositis-specific antibodies
P155/140 (Anti-TIF1- γ)	3(33%)	24(35%)	0.736
MJ (Anti-NXP-2)	1 (11%)	10 (14.7%)
Mi2	1 (11%)	3 (4.4%)
MDA5 (anti-CADM140)	2 (22%)	4 (5.8%)
Other MSA	0 (0%)	3 (5.8%)
Multiple MSAs	1 (11%)	10 (14.7%)
Negative	1 (11%)	13 (19%)

2.2. Disease parameters

For all patients included in the study, we obtained their available laboratory evaluation and clinical evaluations at the time of JDM diagnosis that were included in the Cure JM Biorepository and registry. Clinical data obtained included age at visit of diagnosis, CMAS, and disease activity score (DAS) including the skin DAS (sDAS) and muscle DAS (mDAS). The DAS is a validated disease activity score based on both clinical evidence of muscle weakness and clinical assessment of the patient's skin rash [9]. The CMAS is validated strength tool and was independently verified by a physical therapist [10]. Nailfold capillary end row loops (ERL) were assessed by averaging the number of end row capillary loops per mm in the eight digits, excluding thumbs, and was obtained at either diagnosis or first follow up visit [11].

The laboratory evaluation included CRP, vWF antigen, neopterin, lymphocyte cell flow cytometry results, C3, C4, IgG, ANA titer and muscle enzymes including creatine phosphokinase (CK), lactate dehydrogenase (LDH), alanine aminotransferase (ALT), aspartate aminotransferase (AST), and aldolase. MSA was measured by immunoprecipitation and immunodiffusion by the Oklahoma Medical Research Foundation.

2.3. JDM case counts

The number of cases of JDM were established using the Cure JM biorepository. The number of JDM patients included in the registry were used to provide the number of JDM patient per year prior the pandemic. For the post- pandemic group, the JM registry was also used to determine the number of new cases of JDM for the year. Research participation and registry inclusion was hampered during the pandemic and clinical and laboratory data for patients not enrolled in the registry was not included or obtained for this study. The number of cases per 1 year period were confirmed using Epic SlicerDicer (an analytic tool built into the Epic EMR) with data shown in Fig. 1 .

Fig. 1.

Shows the number of new JDM cases per year.

2.4. Statistical analysis

The statistical software R (version 4.2.1) was used to perform Welch's t-test to compare the baseline characteristics, laboratory parameters, and clinical characteristics between groups. Chi Square testing was used to analyze the racial backgrounds and MSA differences between the groups. GraphPad Prism 9.4.1 and Microsoft Office were used to generate the figures.

3. Results

As seen in Table 1, there were no significant differences in age at first visit or racial background. There were no statistically significant differences between the MSAs in the pre- pandemic and post- pandemic group (Chi Sq. p = 0.736).

The mean NK Cell absolute number for the post- pandemic group was significantly lower than in the pre- pandemic group (90.75/mm3±76 vs 163/mm3±120, P = 0.038). This was mirrored by the NK percentage with the post- pandemic group having significantly lower NK cell percentage than the pre- pandemic group (3.63%±2.3 vs. 6.6% ±4.1, P = 0.008). There were no statistically significant differences in DAS, mDAS, sDAS, CD3 positive cells (count or percent), CD19 positive cells (count or percent), vWF antigen, Neopterin, CMAS, CPK, LDH, Aldolase, ALT, AST, CRP, immunoglobulin levels, C3, C4, ERLs, vitamin D level or ANA titer, as demonstrated in Table 2 . Of note, the total number of new JDM cases rose from an average of 6.9 cases per year prior to January 2020 to an average 9 cases per year from January 1st 2020 to December 31st 2021, as seen in Fig. 1.

Table 2.

Pre and Post COVID group: Disease Activity Markers, Clinical and Laboratory Results.

Flow Cytometry (cells/ul)	Reference Range	Post-Pandemic	Pre-Pandemic	p-Value
Total T cells (CD3+)	895–8192	1393 ± 559	1694 ± 1075	0.24
T helper cells (CD3+ CD4+)	488–4552	887 ± 381	1109 ± 773	0.20
T cytotoxic cells (CD3+ CD8+)	270–3749	456 ± 187	513 ± 341	0.48
NK cells (CD16+/CD56+)	121–1581	90.75 ± 76	163 ± 120	0.038
B cells (CD19+)	203–2711	830 ± 633	772 ± 460	0.81
Laboratory Findings
C Reactive Protein (mg/dl)	0–0.8	0.2 ± 0.11	0.256 ± 0.342	0.46
Neopterin (nmol/l)	0–10	28.14 ± 13.87	20.19 ± 13.04	0.19
von Willebrand factor antigen (%)	36–241	188.14 ± 62.07	158.22 ± 80.78	0.28
Immunoglobulin G (mg/dl)	331–1509	898.57 ± 212	1024.08 ± 359	0.26
Complement Component 3 (mg/dl)	86–184	98.58 ± 26.45	95.67 ± 17.92	0.84
Complement Component 4 (mg/dl)	20–59	21.45 ± 5.02	19.15 ± 4.93	0.56
Hydroxy-25 Vitamin D (ng/ml)	30–100	26.49 ± 11.84	30.65 ± 14.28	0.40
Anti-Nuclear Antibody (ANA) Titer	<80	1440 ± 1097	1220.5 ± 962.73	0.69
Muscle enzymes
Creatine phosphokinase (IU/L)	26–279	1543 ± 1942	2486 ± 8726	0.47
Aspartate aminotransferase (IU/L)	18–65	158.60 ± 155.61	134.61 ± 264.98	0.71
Lactate dehydrogenase (IU/L)	147–438	518.22 ± 136.39	475.98 ± 335.94	0.51
Aldolase (U/L)	3.4–11.8	19.99 ± 9.63	19.15 ± 28.26	0.87
Clinical Findings	Testing Range	Post-Pandemic	Pre-Pandemic	p-Value
Disease activity score—total	0–20	11.39 ± 2.98	10.82 ± 3.29	0.60
Disease activity score—skin	0–9	6.00 ± 1.80	5.18 ± 2.04	0.23
Disease activity score—muscle	0–11	5.39 ± 2.40	5.48 ± 2.83	0.92
Childhood Myositis Assessment Scale	0–52	33.67 ± 16.03	31.39 ± 14.69	0.75
Nailfold capillary end row loops	Normal ≥ 7	5.22± 1.24	5.03 ± 1.47	0.68

4. Discussion

COVID-19 is novel single stranded RNA virus with emerging associations with autoimmunity. Our study showed a moderate increase in JDM cases during the pandemic which is consistent with other studies that show JDM increased during the COVID pandemic [12]. The temporal increase of JDM during the COVID pandemic further bolsters the hypothesis that RNA viruses may trigger JDM as suggested by a gene expression study of untreated JDM muscle biopsies from the cure JM biorepository supporting a viral RNA antigen trigger for JDM [13]. There were no significant changes in JDM phenotype with respect to either clinical or laboratory features beyond the lower NK cell counts in the post-pandemic group. Prior research as shown that COVID-19 causes changes in nailfold capillaries in both acute COVID and MIS-C [14]. We did not find a change in nailfold capillary number between the post and pre-pandemic groups, but both groups did have decreased density of nailfold capillary end row loops which is a vascular hallmark of active JDM. Although there were no readily apparent changes in the JDM phenotype, this study was limited by a small sample size after the COVID-19 pandemic and may not have detected subtle changes. Furthermore, among the newly diagnosed JDM patients after the COVID-19 pandemic it is difficult to ascertain whether their JDM was triggered by COVID-19 or by another mechanism. This is further complicated by changes in the prevalence of COVID-19 variants, COVID19 mRNA vaccination, and the unknown effect this may have had on any unique post-pandemic JDM phenotype.

NK cells have long been known to be impaired in JDM and low NK cells have been associated with higher disease activity scores and JDM severity [15,16]. Despite this, we did not find the post-pandemic JDM group to have higher disease activity scores or any significant change in disease severity. Interestingly, both of the patients with known antecedent COVID-19 infections had lower NK cells (1% and 3% respectively) than the average NK cell percentage in both pre-pandemic (3.6%) and post-pandemic (6.6%) groups. It may be that low NK cells in the post-pandemic group and especially among the two patients with antecedent COVID-19 infections represent JDM activity, but alternatively lower NK cells may reflect a direct viral cause in some cases. It has been observed that patients with severe COVID-19 infection experienced a significant decrease in the number of NK cells, accompanied by functional impairment such as increased expression of the inhibitory molecule NKG2A [18]. Viral illnesses have been shown to unearth or trigger unique phenotypes in association with NK cell dysfunction. For example, a postmortem genetic analysis revealed mutations in PRF1 and LYST of 16 patients with fatal outcomes due to HLH (Hemophagocytic lymphohistiocytosis) triggered by H1N1 influenza pandemic [17]. Low peripheral NK cells in JDM are hypothesized to be a consequence of migration into peripheral muscles. Here we speculate that the lower-than-expected NK cells in the post-pandemic group may reflect not only JDM disease activity, but possibly an intrinsic defect in NK cells leading to an aberrant response to COVID-19.

The limitations of this study include the following: First, the study was not powered to find subtle changes or changes in the MSA distribution in the post-pandemic group due to a small sample size. Second, the study did not clearly distinguish between patients with JDM triggered by COVID versus another mechanism. This is challenging due to asymptotic infections, the lack of antibody responses in some patients, IVIG use making antibody tests less reliable, and now widespread vaccination further complicating the use of antibody responses to ascertain antecedent COVID-19 infection. Given the rarity of JDM, this study remains clinically useful showing that on the whole the JDM phenotype remains unchanged both in the larger retrospective study but also in the case reports beyond lower NK cells at diagnosis and increase in the overall JDM rate after the COVID-19 pandemic. Finally, full enrollment into the registry and biorepository was hampered, especially during the initial phase of the COVID-19 pandemic, which may have led to unincluded data for post-pandemic JDM patients.

5. Conclusion

In conclusion, we found that JDM cases increased during the pandemic. The JDM phenotype prior to and after the pandemic onset overall appears unchanged beyond lower NK cell count in the post-pandemic group. The lower NK cells found in the post-pandemic group and especially among the two patients who reported an antecedent COVID-19 infection highlights NK cells as a possible future research topic for virally triggered JDM. Further research into NK cell dysfunction in JDM may elucidate the mechanisms utilized by viruses to trigger JDM.

Funding

This work is supported by the NIH/ National Institute of Arthritis and Musculoskeletal and Skin Diseases[Grant Number T32 AR007611–13]; The Brinson Foundation; the CureJM Foundation as well as individual donors; and NUCATS Clinical and Translational Science Award (CTSA) grant from the National Institutes of Health (NIH)[Grant Number UL1TR001422]

Author contributions

Christopher Costin, MD, Rheumatology Fellow, Author, performed data analysis. Amer Khojah, MD, Immunology/Rheumatology Attending, review and NK cell expertise; GA Morgan, MA, CCRP provided the clinical data from the JDM Registry. Lauren M. Pachman, MD, Professor of Pediatrics, instigated the data collection, data entry and provided her lab for processing. Marisa Klein-Gitelman, MD provided the clinical observations and obtained the laboratory data for the children with JDM, manuscipt review.

Declaration of competing Interest

Lauren Pachman reports financial support was provided by Cure JM Foundation. Dr. Klein reports a relationship with Pfizer that includes: funding grants. Dr. Klein reports a relationship with National Institute of Arthritis and Musculoskeletal and Skin Diseases that includes: funding grants. Dr. Klein reports a relationship with AbbVie Inc that includes: funding grants. Dr. Klein reports a relationship with Lupus Foundation of America Inc that includes: funding grants. Dr. Klein reports a relationship with Abbott Laboratories that includes: funding grants. Dr. Klein reports a relationship with Bristol Myers Squibb Co that includes: funding grants. Author for UpToDate- Dr. Klein

Data availability

• The authors do not have permission to share data.