Patients’ characteristics
One hundred fifty-two IIM patients (DM, n = 50; ASyS, n = 46; IMNM, n = 32 and IBM, n = 24) and 33 healthy donors were included. Main patients’ characteristics are shown in Table 1.
Table I: Patients’ characteristics at blood sampling timepoint
Diagnosis | DM | ASyS | IMNM | IBM | Total | p |
n (%) | 50 (33) | 46 (30) | 32 (21) | 24 (16) | 152 (100) | - |
Age (year) | 53.2 ± 15.4 | 48.6 ± 14.5 | 49.6 ± 19.1 | 69.3 ± 8.3 | 53.6 ± 16.6 | p < 0.001 |
MSA, n (%) | 35 (70) | 46 (100) | 32 (100) | 12 (100) | 125 | - |
MAA, n (%) | 13 (26) | 37 (80) | 9 (28) | 3 (13) | 62 | - |
MMT8 (0-150) | 142 [126–150] | 150 [132–150] | 132 [115–146] | 120 [94–133] | 138 [119–150] | p < 0.001 |
CK level (UI/ml) | 112 [60–460] | 550 [123–1500] | 780 [249–1332] | 586 [296–1123] | 432 [109–974] | P < 0.001 |
MDAAT (0–60) | 10 [3.5–17] | 9 [2-21.5] | na | na | 10 [2.5–17] | 0.65 |
PGA (0–10) | 5 [2–8] | 5 [2–8] | 5 [2–7] | na | 5 [2–8] | 0.85 |
Corticosteroids n (%) | 33 (66) | 25 (54) | 22 (68) | 0 | 80 (52) | 0.36 |
Immunomodulator, n (%) | 23 (46) | 26 (57) | 17 (53) | 0 | 66 (43) | 0.58 |
MMT8: manual muscle testing 8, CK: creatine phosphokinase, MDAAT: myositis disease activity assessment tool, na: non available, PGA: physician global assessment. Immunomodulator: methotrexate, azathioprine, ciclosporine, rituximab, cyclophosphamide, hydroxychloroquine. MSA: myositis specific antibody (anti-Mi2, -Tif-1γ, -NXP2, -MDA5, -SAE1, -Jo1, -PL7, -PL12, -HMGCoa, -SRP, -cn1a). MAA: myositis-associated antibody (anti-Ro52, -Ro60, RF, -CCP, -RNP, -DNA). DM: Dermatomyositis, ASyS: Anti-synthetase syndrome, IMNM: Immune mediated necrotizing myopathie, IBM: Inclusion body myositis.
As expected, IBM patients were older and displayed a lower MMT8 score compared to DM and ASyS. MSA were detected in 70.6% of DM patients (anti-Mi2, n = 10; -TIF1γ, n = 12; -NXP2, n = 7; -MDA5, n = 5 and -SAE, n = 2), by definition all ASyS were antibody positive (anti-Jo1, n = 39; -PL7, n = 4; -PL12, n = 2 and –EJ, n = 1), and all IMNM patients were seropositive (anti-SRP, n = 13, -HMGCR, n = 19).
No difference was observed in the therapeutic profile, including the use of corticosteroids and immunosuppressors, between IMNM, DM, and ASyS while IBM patients did not receive any treatment.
Increased levels of type I and II IFNs depend on the myositis subgroups
Serum IFN-α level was significantly higher in DM (0.07 [0.03–0.23] pg/ml) and ASyS (0.07 [0.02–0.16] pg/ml) compared to HD (0.02 [0.01–0.05] pg/ml; p < 0.005 and p < 0.05 respectively) whereas it was not significantly different in IMNM (0.03 [0.01–0.09] pg/ml) or IBM (0.02 [0.02–0.03] pg/ml) compared to HD (Fig. 1a). One quarter (26%; n = 13/50) of DM and 20% ASyS patients (n = 9/46) had increased IFN-α level above the positivity threshold while only 3% of IMNM (n = 1/32) and 4% of IBM patients (n = 1/24) had increased levels.
Only DM patients had significantly higher IFN-β level (1.24 [1.24–6.31] pg/ml) compared to HD (1.24 [1.24–1.24] pg/ml, p < 0.005) (Fig. 1b). IFN-β was increased in 34% (n = 17/50) of DM patients, and 12% of IBM patients (n = 3/24) while no ASyS or IMNM patients presented an increased level.
IFN-γ level was significantly increased in all IIM subgroups (ASyS (1.05 [0.47–2.46] pg/ml), DM (0.90 [0.55–2.09] pg/ml), IMNM (0.96 [0.42–1.29] pg/ml) and IBM (0.93 [0.42–2.09] pg/ml)) compared with HD (0.46 [0.29–0.59] pg/ml), p < 0.05) (Fig. 1c). One third of ASyS patients (37%; n = 17/46), one quarter of DM patients (26%; n = 13/50) and IBM patients (25%, n = 6/24) and 16% of IMNM patients (n = 5/32) had an increased level of IFN-γ.
IFN levels and disease activity
Type I and II IFN levels are correlated with disease activity of DM and ASyS
Correlation between IFN level and disease activity showed that disease activity was strongly correlated with type-I IFN, IFN-α (r = 0.76 [0.60–0.86], p < 0.001) (Fig. 2a) and IFN-β (r = 0.58 [0.35–0.74], p < 0.01) (Fig. 2c) in DM. A mild correlation with IFN-γ (r = 0.36 [0.05–0.56], p = 0.02) was observed.
ASyS also demonstrated that disease activity correlated with IFN-α (r = 0.55 [0.34–0.76], p < 0.001) (Fig. 2e) and IFN-γ levels (r = 0.46 [0.15–0.66], p = 0.003) (Fig. 2g). Of note, no ASyS patient presenting an active disease had increased IFN-β level.
In IMNM, only IFN-γ level was significantly correlated with disease activity (r = 0.48 [0.14–0.71], p = 0.006) whereas IFN-α (r = 0.23 [-0.14-0.55], p = 0.2) and IFN-β (r=-0.07 [-0.43-0.31], p = 0.7) were not. Of note, correlation between CK levels and disease activity was very high (r = 0.87 [0.73–0.94], p < 0.001) for IMNM patients.
Multivariate analysis including IFN-α, and IFN-γ showed that only IFN-α was associated with active disease in DM patients (OR = 9.5 [3.1–45.9], p < 0.001). Concerning ASyS patients, only IFN-α was statistically associated with disease activity (OR = 5 [1.9–17.9], p = 0.004), and there was a trend for IFN- γ (p = 0.08). No IFN subtype was associated with disease activity in the IMNM subgroup.
Longitudinal analysis
Focusing on DM and ASyS treatment-naive patients at diagnosis, we performed a serial longitudinal analysis (DM, n = 6/11 and ASyS, n = 4/11) (supplementary Fig. 1 and Fig. 3). The majority of DM patients had a high type I IFN level at diagnosis and a decrease parallel to the clinical improvement (Fig. 3a, Fig. 3b and supplementary Fig. 1). IFN-γ level wasn’t associated with the disease activity in follow-up of DM (Fig. 3c). In most ASyS patients increased levels of IFN-α and IFN-γ but not IFN-β were observed at the diagnosis (Supplementary Fig. 2). Similarly, these levels decreased following the clinical improvement.
Sensitivity and specificity of IFNs to discriminate active and inactive DM and ASyS patients
Next, we aimed to define the threshold level of IFN corresponding to active disease. Active DM patients had higher level of IFN-α (0.26 [0.09–0.53] pg/ml) compared to inactive DM patients (0.03 [0.01–0.07] pg/ml, p < 0.001) (Supplementary Fig. 2a). ROC analysis showed that the area under the curve (AUC) was 0.88 (IC95 0.79–0.98; p < 0.001) (Fig. 2b). An IFN-α threshold above 0.11 pg/ml had a 75% sensitivity and 96.2% specificity, and 19.5 positive likelihood ratio to discriminate active from inactive DM patients (Table 2).
Table 2
Sensitivity and specificity of different IFNs as activity biomarkers
| IFN level (pg/ml) | Se (%) | Sp (%) | Likelihood ratio + |
DM IFN-α | 0.11 | 75 | 96 | 19.50 |
IFN-β | 1.6 | 57 | 88 | 4.90 |
IFN-γ | 1.06 | 71 | 77 | 3.07 |
ASyS IFN-α | 0.06 | 82 | 81 | 4.42 |
IFN-γ | 1.05 | 80 | 73 | 2.97 |
DM: Dermatomyositis, ASyS: Anti-synthetase syndrome, IFN: Interferon, Se : sensitivity, Sp : Specificity.
Active DM patients had higher level of IFN-β (4.62 [1.24.-27.96]) compared to inactive DM patients (1.24 [1.24–1.24] pg/ml, p = 0.0001) (Supplementary Fig. 2b) with a 0.75 AUC (IC95 0.61–0.89; p = 0.0027) (Fig. 2d). Of note, IFN-γ levels were higher in active DM patients (1.417 [0.81–2.74] pg/ml) compared to inactive ones (0.64 [0.38–1.20] pg/ml, p = 0.007).
Active ASyS patients had higher IFN-α level (0.16 [0.08–0.36] pg/ml) compared to inactive ASyS patients (0.03 [0.01–0.06] pg/ml, p < 0.001) (Supplementary Fig. 2c). The sensitivity was 82.3% and the specificity 80.8% for an IFN-α threshold above 0.06 pg/ml (AUC = 0.86, IC95(0.74–0.97); p < 0.001) (Fig. 2f). Active ASyS patients had higher level of IFN-γ (2.28 [1.18–3.26] pg/ml) compared to inactive ASyS patients (0.82 [0.34–1.33] pg/ml, p = 0.004) (Supplementary Fig. 2d) and sensitivity and specificity were 80% and 73.1% respectively at a threshold above 1.05 pg/ml (Table 2) (AUC = 0.75, IC95(0.60–0.90); p = 0.003 (Fig. 2 h)).