Abstract
Objective:
Novel autoantibody specificities including anti-CCAR1 were recently discovered in adult patients with anti-TIF1γ−positive dermatomyositis (DM) and were associated with attenuated cancer emergence. The aims of the present study were to examine whether these autoantibodies occur in patients with juvenile-onset dermatomyositis (JDM) and to determine their associated features.
Methods:
Sera from 150 patients with anti-TIF1γ autoantibody-positive JDM in a cross-sectional cohort and 90 juvenile healthy controls were assayed for anti-CCAR1, anti-C1Z1, anti-IMMT, anti-TBL1XR1, and anti-Sp4 autoantibodies. Demographics, myositis autoantibodies, clinical features, medications, outcomes, and HLA-DRB1 and HLA-DQA1 alleles were compared between those with and without these autoantibodies.
Results:
Any one of the anti-TIF1γ-associated autoantibodies was present in 44 (29%) patients overall, including 25 (17%) with anti-Sp4, 22 (15%) with anti-TBL1XR1, 14 (9%) with anti-CCAR1, 2 (1%) with anti-C1Z1, and 2 (1%) with anti-IMMT autoantibodies. These anti-TIF1γ-associated autoantibodies frequently co-occurred. Patients with any of the anti-TIF1γ-associated autoantibodies had less frequent falling (34% [15] vs 53% [56], p=0.032) and lower peak muscle enzymes. None of the patients had cancer. Among White patients, HLA-DRB1*03 was protective against an anti-TIF1γ-associated autoantibody (OR 0.20, 95% CI 0.07–0.52).
Conclusion:
Autoantibodies associated with anti-TIF1γ were found in isolation and in combination among a subset of patients with JDM. Patients with these autoantibodies had less severe muscle disease and were not enriched for HLA-DRB1*03. Additional autoantibodies among anti-TIF1γ-positive patients with JDM likely contribute to the heterogeneity of the anti-TIF1γ serologic subgroup.
Keywords: juvenile idiopathic inflammatory myopathies, myositis autoantibodies, anti-TIF1γ
Anti-transcriptional intermediary factor 1γ (TIF1γ) autoantibodies are among the most common specificities in both adult and juvenile patients with dermatomyositis (DM, JDM) (1, 2). Anti-TIF1γ autoantibodies are often characterized by widespread, severe cutaneous manifestations including photosensitive rashes, skin ulceration, and lipodystrophy (2, 3). In JDM, anti-TIF1γ autoantibodies are associated with a spectrum of muscle involvement ranging from clinically amyopathic disease to severe weakness, as well as a chronic disease course and longer time to remission (2–6). Importantly, in adult DM anti-TIF1γ autoantibodies are associated with increased risk of cancer (1).
Additional autoantibody specificities enriched in patients with anti-TIF1γ autoantibodies, including anti-CCAR1 and anti-Sp4, were recently discovered in adults with DM (7, 8). These were associated with attenuated cancer emergence and a decrease in cancer risk relative to the general population. In the study of anti-CCAR1 autoantibodies, the likelihood of cancer was found to diminish with greater numbers of additional autoantibody specificities present, suggesting that a broader immune response is protective against cancer emergence (7). Anti-CCAR1 autoantibodies, in a subsequent study of adult DM, were associated with lower peak creatinine kinase as well as lower frequency of elevated creatinine kinase and cutaneous ulceration (9). Interestingly, although cancer is not associated with JDM, anti-Sp4 autoantibodies were found in a cohort of patients with juvenile myositis, almost entirely among those with anti-TIF1γ autoantibodies (10). Anti-Sp4 autoantibodies in juvenile-onset myositis were associated with Raynaud phenomenon and less severe muscle disease, similar to adults, as well as with distinct immunogenetic risk factors.
The aim of the present study was to determine whether additional co-existing autoantibody specificities may further define subgroups among anti-TIF1γ autoantibody-positive patients with JDM. Here, a cross-sectional cohort of patients with anti-TIF1γ autoantibody-positive JDM were screened for additional specificities among those previously associated with anti-TIF1γ autoantibodies in adult DM. The clinical characteristics and immunogenetic risk factors for these autoantibodies were explored.
METHODS
Patients and sera
One hundred fifty patients enrolled between 1989–2022 in National Institutes of Health institutional review board-approved cross-sectional studies (NCT00017914, NCT00055055, NCT00341679, NCT01276470) of the Childhood Myositis Heterogeneity Collaborative Study Group who met classification of probable or definite JDM per Bohan and Peter criteria and were confirmed to have anti-TIF1γ autoantibodies by ELISA (Medical & Biological Laboratories CO., LTD., Nagoya, Aichi, Japan) were included (11, 12). Informed consent was provided by all subjects.
Myositis clinical subgroups included juvenile dermatomyositis (JDM) and juvenile connective tissue myositis (JCTM), which included patients with JDM overlapping with another connective tissue or autoimmune disease; none of the patients had juvenile polymyositis. Demographic features, cumulative disease manifestations, early organ system scores (a continuous measure with values ranging from 0–1) (4), peak serum muscle enzymes, medications received, and outcomes ever achieved including complete clinical response and remission (6) were determined via physician questionnaire and completed by medical record review and patient interviews (13).
All patients (77% female, median age at enrollment 10.9 years (IQR 7.4–15.3)) and 90 juvenile healthy control subjects (66% female, median age at enrollment 10.8 years (IQR 9.2–13.4)) were screened for anti-CCAR1 (cell division cycle and apoptosis regulator protein 1), anti-C1Z1 (Cip1-interacting zing finger protein), anti-IMMT (MICOS complex subunit MIC60), and anti-TBL1XR1 (F-box-like/WD repeat–containing protein TBL1XR1) from samples collected at enrollment. Anti-CCAR1 autoantibodies were assayed by ELISA (9). The same anti-CCAR1–positive serum was included as a reference in every ELISA, and all absorbances were calibrated relative to the reference serum. A cutoff of 0.3244 OD units, which was equal to four standard deviations above the mean value of 90 healthy juvenile control subjects, was considered positive; those with borderline results were considered negative. Anti-C1Z1, anti-IMMT, and anti-TBL1XR1 autoantibodies were assayed by immunoprecipitation (IP) of 35S-methionine-labeled proteins generated by in vitro transcription and translation (7). Ninety patients from the present cohort and 91 juvenile healthy control subjects were previously screened for anti-Sp4 (specificity protein 4) autoantibodies by ELISA; the remaining 60 patients from the present cohort were assayed using the same method (10). Other myositis autoantibodies were tested using validated IP, immunoblotting (IB), IP-IB, and ELISA methods (10, 13).
Genotyping for HLA-DRB1 and HLA-DQA1 alleles was performed in White patients, and those with both alleles available were included in analyses (14). Parental HLA typing from the National Institute of Environmental Health Sciences Study of Families with Twins or Siblings Discordant for Rheumatic Disorders (NCT00055055) was available for 50 patients.
Analyses
Categorical and continuous variables are reported as percentage (frequency) and median (interquartile range), respectively. Missing data, where applicable, are reported. Chi-squared, Fisher exact test, or Wilcoxon rank-sum test were used for comparisons between groups. Logistic regression was used to present odds ratios (OR) with 95% confidence intervals (CI) for statistically significant differences in HLA alleles. A two-sided p<0.05 without adjustment was considered significant in this exploratory study; for HLA analyses, where the exact number of tests was known, the Benjamini-Hochberg procedure was applied for the group of alleles included in each analysis and corrected p-values (Pcorr) are reported. Analyses were performed using R Statistical Software (Version 4.1.0; R Core Team 2021) (15).
RESULTS
Anti-TIF1γ-associated autoantibodies are present and frequently co-occur in patients with JDM
Anti-CCAR1 autoantibodies were present in the sera of 14 (9%) patients with anti-TIF1γ-positive JDM and 1 (1%) healthy control subject (Supplemental Figure 1). Anti-TBL1XR1 autoantibodies were present in 22 (15%), anti-C1Z1 in 2 (1%), and anti-IMMT in 2 (1%) patients and none of the controls. Additionally, anti-Sp4 autoantibodies were present in 25 (17%) patients and none of the controls. Forty-four (29%) patients overall had at least one of these five autoantibodies. Anti-TIF1γ autoantibody titer was lower among patients with one of these additional autoantibodies compared to those without (35.0 [15.5, 67.3] vs 56.0 [26.0, 111.8], p=0.022).
Anti-TIF1γ-associated autoantibodies occurred in 7 different combinations, and anti-C1Z1 autoantibodies were the only autoantibody not to occur in isolation (Figure 1). Anti-CCAR1 autoantibodies more often co-occurred with other autoantibodies (71%, 10 of 14) than in isolation, whereas anti-TBL1XR1 and anti-Sp4 autoantibodies occurred approximately as often in isolation as in combination.
Figure 1.
Combinatorial expression of autoantibodies in anti-TIF1γ autoantibody-positive juvenile dermatomyositis patient sera.
Horizontal black bars show the counts by individual autoantibody specificity. Vertical black bars show the number of patients with each autoantibody combination in the cohort. In the matrix, each column represents one autoantibody combination. Gray circles denote absence of a specific autoantibody, black circles denote presence of a specific autoantibody, and connecting black lines denote multiple specificities present in a combination.
Anti-TIF1γ-associated autoantibodies are associated with less severe muscle involvement
Patients with any of the anti-TIF1γ-associated autoantibodies were more often female (91% [40] vs 71% [75], p=0.008), but other demographic features did not differ compared to patients without one of these autoantibodies (Table 1). The frequency of clinical subgroups and myositis-associated autoantibodies between patients with and without any anti-TIF1γ-associated autoantibody did not differ. All patients with an anti-TIF1γ-associated autoantibody had JDM. None of the patients in this cohort had cancer.
Table 1.
Demographics and myositis autoantibodies of anti-TIF1γ autoantibody-positive juvenile dermatomyositis patients with and without anti-TIF1γ associated autoantibodies
| Anti-TIF1γ-associated Ab positive, N = 44 | Anti-TIF1γ-associated Ab negative, N = 106 | p-value | |
|---|---|---|---|
| Demographics | |||
| Female | 91% (40 / 44) | 71% (75 / 106) | 0.008 |
| Race/Ethnicity | |||
| White | 86% (38 / 44) | 74% (78 / 106) | 0.089 |
| Black | 2% (1 / 44) | 6% (6 / 106) | 0.7 |
| Hispanic | 5% (2 / 44) | 7% (7 / 106) | >0.9 |
| Other | 7% (3 / 44) | 14% (15 / 106) | 0.2 |
| Age at diagnosis (years) | 6.9 (4.2, 11.0) | 7.4 (5.2, 10.6) | 0.4 |
| Delay in diagnosis (months) | 5.7 (2.9, 11.1) | 4.0 (2.0, 8.6) | 0.083 |
| Duration from diagnosis to enrollment (years) | 2.0 (0.7, 4.5) | 1.9 (0.7, 3.9) | 0.8 |
| Unknown | 1 | 9 | |
| Duration from diagnosis to last follow up (years) | 3.9 (1.4, 7.8) | 4.8 (2.3, 9.4) | 0.3 |
| Unknown | 1 | 9 | |
| Clinical subgroup | |||
| JDM | 100% (44 / 44) | 98% (104 / 106) | >0.9 |
| JCTM | 0% (0 / 44) | 2% (2 / 106) | >0.9 |
| Myositis autoantibodies | |||
| Anti-Ro52 | 10% (4 / 40) | 12% (11 / 89) | 0.8 |
| Anti-Ro60 | 7% (3 / 44) | 12% (13 / 106) | 0.4 |
| Anti-NT5C1A | 37% (11 / 30) | 26% (20 / 76) | 0.3 |
| Anti-Sp4 | 57% (25 / 44) | 0% (0 / 106) | <0.001 |
| Anti-TBL1XR1 | 50% (22 / 44) | 0% (0 / 106) | <0.001 |
| Anti-CCAR1 | 32% (14 / 44) | 0% (0 / 106) | <0.001 |
| Anti-IMMT | 5% (2 / 44) | 0% (0 / 106) | 0.085 |
| Anti-C1Z1 | 5% (2 / 44) | 0% (0 / 106) | 0.085 |
Values are % (n / N) or median (IQR). Other myositis-specific autoantibodies were found in 10 patients, including 1 with anti-HMGCR (negative for TIF1γ-associated Abs) and 9 with anti-Mi2 (1 was positive for both anti-CCAR1 and anti-TBL1XR1 autoantibodies). Both patients with JCTM had JDM, one with overlapping systemic sclerosis and one with overlapping Crohn disease.
Abbreviations: Ab, autoantibody; JCTM, juvenile connective tissue myositis; JDM, juvenile dermatomyositis; MAA, myositis-associated autoantibody; NT5C1A, cytosolic 5’-nucleotidase 1A.
Those with an anti-TIF1γ-associated autoantibody had less severe muscle involvement (Table 2). Falling (34% [15] vs 53% [56], p=0.032) was less common and all peak serum muscle enzyme levels were significantly lower in patients positive for any anti-TIF1γ-associated autoantibody. These patients also had lower early muscle system scores (0.3 vs 0.4, p=0.020).
Table 2.
Clinical features of anti-TIF1γ autoantibody-positive juvenile dermatomyositis patients with and without anti-TIF1γ-associated autoantibodies
| Anti-TIF1γ-associated Ab positive, N = 44 |
Anti-TIF1γ-associated Ab negative, N = 106 |
p-value | |
|---|---|---|---|
| Muscle involvement | |||
| Distal weakness | 37% (16 / 43) | 51% (54 / 105) | 0.12 |
| Muscle atrophy | 39% (17 / 44) | 42% (44 / 105) | 0.7 |
| Falling | 34% (15 / 44) | 53% (56 / 105) | 0.032 |
| Wheelchair use | 7% (3 / 43) | 18% (19 / 105) | 0.084 |
| Cutaneous involvement | |||
| Heliotrope rash | 95% (42 / 44) | 90% (95 / 106) | 0.3 |
| Gottron papules | 98% (43 / 44) | 98% (104 / 106) | >0.9 |
| V-sign rash | 41% (18 / 44) | 50% (52 / 105) | 0.3 |
| Shawl sign rash | 23% (10 / 44) | 38% (39 / 104) | 0.081 |
| Cuticular overgrowth | 53% (23 / 43) | 49% (50 / 103) | 0.6 |
| Periungual capillary dilatation | 91% (40 / 44) | 90% (95 / 105) | >0.9 |
| Raynaud phenomenon | 23% (10 / 44) | 5% (5 / 106) | 0.002 |
| Malar rash | 95% (42 / 44) | 94% (100 / 106) | >0.9 |
| Photosensitivity | 70% (31 / 44) | 70% (72 / 103) | >0.9 |
| Linear extensor erythema | 59% (26 / 44) | 53% (56 / 105) | 0.5 |
| Erythroderma | 16% (7 / 44) | 8% (9 / 106) | 0.2 |
| Skin ulceration | 11% (5 / 44) | 29% (31 / 106) | 0.020 |
| Edema | 11% (5 / 44) | 17% (18 / 106) | 0.4 |
| Calcinosis | 34% (15 / 44) | 32% (34 / 106) | 0.8 |
| Lipodystrophy | 16% (7 / 44) | 16% (17 / 106) | >0.9 |
| Panniculitis | 11% (5 / 44) | 8% (8 / 106) | 0.5 |
| Other involvement | |||
| Arthritis | 41% (18 / 44) | 53% (56 / 106) | 0.2 |
| Dysphagia | 39% (17 / 44) | 49% (52 / 106) | 0.2 |
| Dysphonia | 20% (9 / 44) | 35% (37 / 105) | 0.075 |
| Dyspnea on exertion | 27% (12 / 44) | 22% (23 / 105) | 0.5 |
| Interstitial lung disease | 2% (1 / 44) | 1% (1 / 105) | 0.5 |
| Weight loss | 39% (17 / 44) | 37% (39 / 106) | 0.8 |
| Fever | 36% (16 / 44) | 35% (37 / 106) | 0.9 |
| Peak serum muscle enzymes | |||
| Creatine kinase (IU/L) | 386 (164, 668) | 840 (310, 2,793) | 0.004 |
| Unknown | 5 | 7 | |
| Aldolase (IU/L) | 9.5 (6.8, 12.1) | 13.0 (8.6, 19.9) | 0.002 |
| Unknown | 10 | 19 | |
| AST (IU/L) | 43 (30, 66) | 82 (51, 155) | <0.001 |
| Unknown | 7 | 13 | |
| ALT (IU/L) | 41 (16, 65) | 63 (41, 107) | <0.001 |
| Unknown | 10 | 22 | |
| LDH (IU/L) | 299 (222, 415) | 427 (334, 681) | <0.001 |
| Unknown | 9 | 30 | |
| Early system symptom scores | |||
| Muscle score | 0.3 (0.2, 0.4) | 0.4 (0.2, 0.6) | 0.020 |
| Unknown | 0 | 4 | |
| Cutaneous score | 0.4 (0.3, 0.5) | 0.4 (0.3, 0.5) | 0.7 |
| Unknown | 1 | 1 |
Values are % (n / N) or median (IQR).
Abbreviations: ALT, alanine transaminase; AST, aspartate transaminase; IU/L, international units per liter; LDH, lactate dehydrogenase.
Additionally, in patients with an anti-TIF1γ-associated autoantibody Raynaud phenomenon was more frequent (23% [10] vs 5% [5], p=0.002) while skin ulceration was less common (11% [5] vs 29% [31], p=0.020). The frequency of other cutaneous manifestations did not differ. There were no differences in medications received or outcomes except for remission, which was less commonly achieved in patients with an anti-TIF1γ-associated autoantibody (5% [2] vs 21% [19], p=0.024) (Supplemental Table 1).
Besides anti-Sp4 autoantibodies, which have already been described in juvenile-onset myositis, anti-TBL1XR1 and anti-CCAR1 autoantibodies were the most frequent anti-TIF1γ-associated autoantibodies (Figure 1) (10). The clinical associations of these autoantibodies were similar. Anti-TBL1XR1 autoantibodies were frequently associated with co-occurring anti-Sp4 (36% [8] vs 13% [17], p=0.013) and anti-CCAR1 (32% [7] vs 5% [7], p=0.001) autoantibodies (Supplemental Table 2). Anti-CCAR1 autoantibodies were also frequently associated with anti-Sp4 (50% [7] vs 13% [18], p=0.002) and anti-TBL1XR1 (50% [7] vs 11% [15], p=0.001) autoantibodies. Patients with these autoantibodies also had less prominent muscle involvement, including lower peak serum muscle enzyme levels or early muscle system scores as well as lower frequency of distal weakness or falling (Supplemental Tables 3). Anti-TBL1XR1 autoantibodies were associated with less frequent skin ulceration (5% [1] vs 27% [35], p=0.021) and anti-CCAR1 autoantibodies with less frequent V-sign rash (21% [3] vs 50% [67], p=0.044) compared to patients with anti-TIF1γ-positive JDM without these autoantibodies. Additionally, Raynaud phenomenon (29% [4] vs 8% [11], p=0.036) and panniculitis (29% [4] vs 7% [9] p=0.021) were more common among those with anti-CCAR1 autoantibodies; each of the four patients with Raynaud phenomenon had co-occurring anti-Sp4 autoantibodies. There were no differences in medications received or outcomes among patients with anti-TBL1XR1 or anti-CCAR1 autoantibodies (Supplemental Table 4).
HLA-DQA1*03 but not HLA-DRB1*03 is an immunogenetic risk factor for anti-TIF1γ-associated autoantibodies
Among White patients with anti-TIF1γ autoantibody-positive JDM, HLA-DRB1*03 was protective against an anti-TIF1γ-associated autoantibody (23% [7] vs 60% [39], Pcorr=0.002; OR 0.20, 95% CI 0.07–0.52). The frequency of HLA-DQA1*03 did not differ between patients with and without one of these additional autoantibodies (37% [11] vs 35% [23], Pcorr>0.9). For patients with anti-TBL1XR1 autoantibodies, which was the most frequent positive autoantibody aside from anti-Sp4 autoantibodies, HLA-DQA1*03 (OR 5.87, 95% CI 1.73–26.7) and HLA-DRB1*04 (OR 6.69, 95% CI 1.97–30.4) were found to be immunogenetic risk factors (Supplemental Table 5). All 9 anti-TBL1XR1 autoantibody-positive patients with HLA-DQA1*03 had HLA-DRB1*04 and vice versa. Further, of the 6 patients with parental haplotypes available, each had at least one copy of the HLA-DRB1*04-DQA1*03-DQB1*03 haplotype.
DISCUSSION
Autoantibodies associated with anti-TIFγ that have previously been reported to be protective against cancer in adults with DM were also found in patients with anti-TIFγ autoantibody-positive JDM. In this carefully phenotyped large cohort, these additional autoantibodies were associated with less prominent muscle involvement and less frequent cutaneous ulceration, likely contributing to the heterogeneity among the anti-TIF1γ autoantibody JDM subgroup.
Among the autoantibodies enriched with anti-TIF1γ, the prevalence of anti-TBL1XR1 autoantibodies was similar in juvenile and adult patients (7). In contrast, autoantibodies against C1Z1, CCAR1 and IMMT were found less frequently in this juvenile cohort. Consistent with findings in adults with DM (7), these autoantibodies were often present in various combinations. Patients with JDM who were positive for any of the anti-TIF1γ-associated autoantibodies had signs of less severe muscle inflammation as well as a lower frequency of cutaneous ulceration, similar to anti-CCAR1 autoantibody-positive adult patients with DM (9). The reason for lower anti-TIF1γ antibody titers and milder muscle disease in patients with anti-TIF1γ-associated autoantibodies is not clear. One possible explanation may be that in this patient subset, the anti-TIF1γ autoantibody titers appear diminished because the other autoantibodies target antigens that may exist complexed with TIF1γ. Consistent with this, co-immunoprecipitation studies indicate that CCAR1 likely exists in a complex with TIF1γ (7). Consequently, the potential myotoxic properties of anti-TIF1γ autoantibodies may be attenuated. Another scenario may be that if tumors occur in juvenile patients, an anti-TIF1γ response is elicited, with or without the associated autoantibodies. Tumors are hypothesized to be eradicated more efficiently in those who mount an immune response against both TIF1γ and the associated autoantigens. Because the tumors are eradicated more quickly in those with anti-TIF1γ and associated autoantibodies, there is less opportunity to develop high affinity anti-TIF1γ autoantibodies. Consequently, there may be lower anti-TIF1γ titers and less severe muscle disease. These autoantibodies, additionally, were associated with few distinct features, including anti-Sp4 autoantibodies with Raynaud phenomenon and anti-CCAR1 autoantibodies with panniculitis (10). Lastly, there were no differences in treatments received or outcomes achieved, aside from a lower rate of remission in patients with any anti-TIF1γ-associated autoantibody.
It is noteworthy that findings reported in separate studies show that autoantibodies against both Sp4 and CCAR1 are associated with a decreased risk of cancer in adults with anti-TIF1γ-positive DM (7, 8). The present study establishes that anti-Sp4 and anti-CCAR1 autoantibodies can co-occur in patients with JDM. Whether these autoantibodies are also present together in adults and may be additionally informative vis-à-vis cancer risk warrants investigation. It has been demonstrated that CCAR1 and TIF1γ are present in a complex, suggesting that anti-CCAR1 autoantibodies may develop via epitope spreading (7). While it is not yet known how these anti-TIF1γ−enriched autoantibodies emerge, their frequent presence in combination suggests a common mechanism. It also remains unclear why these autoantibodies are found in patients with JDM despite the exceptionally low incidence of cancer within this subgroup of myositis. Perhaps subclinical tumors emerge in pediatric patients and, regardless of the diversity of the immune response, are promptly eliminated by a robust juvenile immune system. The pathogenesis of the same myositis autoantibodies, instead, may differ in adults and juveniles. For instance, anti-HMGCR necrotizing myopathy is associated with statin exposure in adults but not juvenile-onset patients and, moreover, different immunogenetic risk factors have been identified in these two populations (16).
As with anti-Sp4 autoantibodies, White anti-TBL1XR1 autoantibody-positive patients with JDM had distinct immunogenetic risk factors, specifically HLA-DQA1*03 and HLA-DRB1*04, likely as a haplotype. HLA-DQA1*03 was recognized as a risk factor for anti-TIF1γ among White patients with DM and JDM in the initial description of these autoantibodies (17). Moreover, HLA-DQA1*03 and HLA-DRB1*04 are risk factors for anti-SAE autoantibodies among Whites with adult DM and the HLA-DRB1*04-DQA1*03-DQB1*03 haplotype has been identified as an immunogenetic risk factor for anti-U1-RNP autoantibodies in Whites with JDM (18, 19). This haplotype has also been implicated in other autoimmune diseases, such as Hashimoto thyroiditis (20).
HLA-DRB1*03 is also an immunogenetic risk factor for anti-TIF1γ autoantibodies in JDM (21). Among White anti-TIF1γ autoantibody-positive patients with JDM in this cohort, however, the presence of this allele was protective against any of the autoantibodies enriched with anti-TIF1γ autoantibodies. The co-occurrence of these autoantibodies, then, may underly the heterogeneity of HLA-DRB1 alleles among anti-TIF1γ autoantibody-positive patients.
This study has several limitations, which are mainly related to the cross-sectional nature of the cohort. Because serial samples were not available, the development and sequence of appearance of anti-TIF1γ-associated autoantibodies could not be examined. Similarly, anti-TIF1γ titers at the time of enrollment, which occurred at different points in disease course, may differ from titers at the time of additional autoantibody emergence or myositis diagnosis. It also is unknown whether anti-TIF1γ autoantibody titers may correlate with quantitative strength assessments, such as Manual Muscle Testing or the Childhood Myositis Assessment Scale, and whether these additional autoantibodies may be found in patients with clinically amyopathic DM. HLA genotyping was not available uniformly, and analysis of alleles was limited to White patients. Sample sizes of patients with each of the different anti-TIF1γ-associated autoantibodies or combinations were small, which did not allow for separate analyses of anti-C1Z1 and anti-IMMT autoantibodies or adjustment for the presence of co-existing autoantibodies. External validation of these findings in other cohorts, especially those less strongly significant, will be required to further understand the prevalence and associated clinical features of these autoantibodies and whether they are present exclusively in anti-TIF1γ-positive patients with juvenile-onset myositis.
In summary, additional autoantibody specificities were identified in patients with anti-TIF1γ autoantibody-positive JDM. These autoantibodies were associated with less severe muscle disease but not the HLA-DRB1*03 immunogenetic risk factor. The presence of these additional autoantibody specificities among anti-TIF1γ autoantibody−positive patients with JDM likely contributes to the heterogeneity of this serologic subgroup.
Supplementary Material
Acknowledgements:
We thank Drs. Hanna Kim and Christopher A. Mecoli for critical review of the manuscript.
Funding:
This research was supported in part by the Intramural Research Programs of the National Institute of Environmental Health Sciences [ZIA ES101074, ZIA ES101081], National Institute of Arthritis and Musculoskeletal and Skin Diseases (NIAMS) [ZIA AR041203], the Clinical Center [ZIC CL002128] of the National Institutes of Health (NIH), NIH/NIAMS R01 AR073208, P30 AR070254, and the Donald B. and Dorothy L. Stabler Foundation.
Data availability statement: All data relevant to the study are included in the article or uploaded as supplementary information.
APPENDIX
Members of the Childhood Myositis Heterogeneity Collaborative Study Group who contributed to this project:
Bita Arabshahi, Susan Ballinger, C April Bingham, John F. Bohnsack, Ruy Carrasco, B Anne Eberhardt, Barbara S. Edelheit, Payam Noroozi Farhadi, Terri H. Finkel, Robert C. Fuhlbrigge, Ellen A. Goldmuntz, Beth S. Gottlieb, Thomas A. Griffin, Ricardo Guirola, William Hannan, Michael Henrickson, Sandy Hong, Olcay Y. Jones, Daniel J. Kingsbury, Takayuki Kishi, Carol B. Lindsley, Gulnara Mamyrova, Paul L. McCarthy, Frederick W. Miller, Simona Nativ, Elif A. Oral, Lauren M. Pachman, Wendy de la Pena, Maria D. Perez, Donald A. Person, Iago Pinal-Fernandez, Sara E. Sabbagh, Kakali Sarkar, Adam Schiffenbauer, Amanda Schlefman, Bracha Shaham, Jennifer Soep, Sangeeta Sule, Ira N. Targoff, Scott A. Vogelgesang, Rita Volochayev, Dawn Wahezi
Footnotes
Ethics approval: This study involved human participants and was approved by the National Institutes of Health Institutional Review board. All subjects provided informed consent.
REFERENCES
- 1.Lundberg IE, Fujimoto M, Vencovsky J, et al. Idiopathic inflammatory myopathies. Nat Rev Dis Primers. 2021;7:86. [DOI] [PubMed] [Google Scholar]
- 2.Rider LG, Shah M, Mamyrova G, et al. The myositis autoantibody phenotypes of the juvenile idiopathic inflammatory myopathies. Medicine (Baltimore). 2013;92:223–43. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Gunawardena H, Wedderburn LR, North J, et al. Clinical associations of autoantibodies to a p155/140 kda doublet protein in juvenile dermatomyositis. Rheumatology (Oxford). 2008;47:324–8. [DOI] [PubMed] [Google Scholar]
- 4.Habers GE, Huber AM, Mamyrova G, et al. Brief report: Association of myositis autoantibodies, clinical features, and environmental exposures at illness onset with disease course in juvenile myositis. Arthritis Rheumatol. 2016;68:761–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Mamyrova G, Kishi T, Targoff IN, et al. Features distinguishing clinically amyopathic juvenile dermatomyositis from juvenile dermatomyositis. Rheumatology (Oxford). 2018;57:1956–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Kishi T, Warren-Hicks W, Bayat N, et al. Corticosteroid discontinuation, complete clinical response and remission in juvenile dermatomyositis. Rheumatology (Oxford). 2021;60:2134–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Fiorentino DF, Mecoli CA, Rosen MC, et al. Immune responses to ccar1 and other dermatomyositis autoantigens are associated with attenuated cancer emergence. J Clin Invest. 2022;132. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Hosono Y, Sie B, Pinal-Fernandez I, et al. Coexisting autoantibodies against transcription factor sp4 are associated with decreased cancer risk in patients with dermatomyositis with anti-tif1gamma autoantibodies. Ann Rheum Dis. 2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Fiorentino D, Mecoli CA, Igusa T, et al. Association of anti-ccar1 autoantibodies with decreased cancer risk relative to the general population in patients with anti-transcriptional intermediary factor 1gamma-positive dermatomyositis. Arthritis Rheumatol. 2023;75:1238–45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Sherman MA, Pak K, Pinal-Fernandez I, et al. Autoantibodies recognizing specificity protein 4 co-occur with anti-transcription intermediary factor 1 and are associated with distinct clinical features and immunogenetic risk factors in juvenile myositis. Arthritis Rheumatol. 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bohan A, Peter JB. Polymyositis and dermatomyositis (first of two parts). N Engl J Med. 1975;292:344–7. [DOI] [PubMed] [Google Scholar]
- 12.Bohan A, Peter JB. Polymyositis and dermatomyositis (second of two parts). N Engl J Med. 1975;292:403–7. [DOI] [PubMed] [Google Scholar]
- 13.Shah M, Mamyrova G, Targoff IN, et al. The clinical phenotypes of the juvenile idiopathic inflammatory myopathies. Medicine (Baltimore). 2013;92:25–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Schiffenbauer A, Faghihi-Kashani S, O’Hanlon TP, et al. The effect of cigarette smoking on the clinical and serological phenotypes of polymyositis and dermatomyositis. Semin Arthritis Rheum. 2018;48:504–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2021. [Google Scholar]
- 16.Kishi T, Rider LG, Pak K, et al. Association of anti-3-hydroxy-3-methylglutaryl-coenzyme a reductase autoantibodies with drb1*07:01 and severe myositis in juvenile myositis patients. Arthritis Care Res (Hoboken). 2017;69:1088–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Targoff IN, Mamyrova G, Trieu EP, et al. A novel autoantibody to a 155-kd protein is associated with dermatomyositis. Arthritis Rheum. 2006;54:3682–9. [DOI] [PubMed] [Google Scholar]
- 18.Betteridge ZE, Gunawardena H, Chinoy H, et al. Clinical and human leucocyte antigen class ii haplotype associations of autoantibodies to small ubiquitin-like modifier enzyme, a dermatomyositis-specific autoantigen target, in uk caucasian adult-onset myositis. Ann Rheum Dis. 2009;68:1621–5. [DOI] [PubMed] [Google Scholar]
- 19.Wedderburn LR, McHugh NJ, Chinoy H, et al. Hla class ii haplotype and autoantibody associations in children with juvenile dermatomyositis and juvenile dermatomyositis-scleroderma overlap. Rheumatology (Oxford). 2007;46:1786–91. [DOI] [PubMed] [Google Scholar]
- 20.Zeitlin AA, Heward JM, Newby PR, et al. Analysis of hla class ii genes in hashimoto’s thyroiditis reveals differences compared to graves’ disease. Genes Immun. 2008;9:358–63. [DOI] [PubMed] [Google Scholar]
- 21.Rothwell S, Chinoy H, Lamb JA, et al. Focused hla analysis in caucasians with myositis identifies significant associations with autoantibody subgroups. Ann Rheum Dis. 2019;78:996–1002. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.