Comparison of clinical parameters between the ADM-NSIP/OP and control groups
The study population consisted of 158 patients with NSIP or OP. Thirty-seven patients met the Sontheimer criteria and were diagnosed with ADM. According to the ATS/ERS consensus criteria, they were separately assigned to NSIP (n=25) and OP (n=15) groups, which included 3 patients with NSIP + OP overlap. When considering the valid scope for NSIP and OP radiological patterns, we included 3 patients in both NSIP and OP ADM groups.
The demographic and main characteristics of NSIP and OP are summarized in Table 1. The mean ages of the ADM-NSIP and ADM-OP groups at diagnosis were 59.48 ± 16.34 and 57.27 ± 12.86 years versus 65.22 ± 14.34 and 62.76 ± 15.16 years in the control-NSIP and control-OP groups, respectively. It seemed that ADM-ILD patients were younger than other ILD patients, but the differences were not significant. Similarly, no difference was observed between groups in regard to sex (p=0.880, p=0.163) and smoking history (p=0.115, p=0.455). Not all of the ADM-ILD patients presented with rashes (52.00%, 46.67%) and systemic symptoms (44.00%, 60.87%), and there were no significant differences between ADM and other ILDs. Overview of the routine blood and serum biochemical indices showed that most of the indicators were within the normal range of fluctuations. However, there were significant differences in alanine aminotransferase (ALT, p=0.009, p=0.018), aspartate aminotransferase (AST, p<0.001, p=0.002), and lactate dehydrogenase (LDH, p<0.001, p=0.001). Furthermore, we found that 40.53% (15/37) and 86.49% (32/37) of ADM-ILD patients had AST and LDH levels above the normal range, respectively, while ALT was within the normal limits (the normal laboratory values of ALT, AST, and LDH were 7-40 U/L, 15-35 U/L, and 106-211 U/L, respectively). Through the comparison between the ADM-OP and control-OP groups, the positive antinuclear antibody level was significantly different (p=0.007). However, considering that most of the control-OP patients had cryptogenic organizing pneumonia (COP), the difference was probably not meaningful. Regrettably, not all ILD patients underwent blood gas analysis and pulmonary function tests, since those results were not documented in the medical records. The number of cases and specific statistical data are reported in Table 1. The subjects with ADM-NSIP had a lower predicted DLco% (56.06 ± 11.32) than the control group, and there was a significant difference in the pulmonary diffusing capacity (p=0.041).
Table 1 Comparison between the ADM-NSIP/OP groups and the corresponding control groups in terms of clinical data.
|
ADM-NSIP (n=25)
|
Control-NSIP (n=92)
|
P
|
ADM-OP (n=15)
|
Control-OP (n=29)
|
P
|
Demographic feature
|
|
|
|
|
|
|
Age (years)
|
59.48 ± 16.34
|
65.22 ± 14.34
|
0.296
|
57.27 ± 12.86
|
62.76 ± 15.16
|
0.471
|
Gender (M/F)
|
8/17
|
28/64
|
0.880
|
6/9
|
18/11
|
0.163
|
Smoking history, n (%)
|
3 (12.00%)
|
25 (27.17%)
|
0.115
|
4 (26.67%)
|
11 (37.93%)
|
0.455
|
Rash, n (%)
|
13 (52.00%)
|
32 (34.78%)
|
0.117
|
7 (46.67%)
|
4 (13.79%)
|
0.028
|
Systemic symptoms, n (%)
|
11 (44.00%)
|
56 (60.87%)
|
0.131
|
5 (33.33%)
|
7 (24.14%)
|
0.722
|
Peripheral blood
|
|
|
|
|
|
|
WBC (×109/L)
|
5.58 ± 2.72
|
7.04 ± 3.43
|
0.982
|
6.01 ± 2.90
|
7.94 ± 2.50
|
0.359
|
Lymphocyte (×109/L)
|
1.04 ± 0.56
|
1.56 ± 0.59
|
0.928
|
1.04 ± 0.74
|
1.52 ± 1.15
|
0.709
|
HB (g/L)
|
119.4 ± 14.65
|
123.49 ± 20.12
|
0.094
|
120.73 ± 19.01
|
124.34 ± 20.01
|
0.907
|
PLT (×109/L)
|
230.76 ± 72.28
|
227.11 ± 85.48
|
0.379
|
236.60 ± 59.46
|
279.66 ± 99.40
|
0.046
|
hsCRP (mg/L)
|
15.81 ± 27.06
|
20.22 ± 34.25
|
0.144
|
19.19 ± 34.61
|
48.08 ± 42.67
|
0.172
|
ESR (mm/h)
|
46.44 ± 37.36
|
41.16 ± 34.69
|
0.706
|
40.93 ± 29.39
|
58.93 ± 35.40
|
0.546
|
ALB (g/L)
|
35.38 ± 5.43
|
37.44 ± 5.06
|
0.534
|
33.55 ± 5.62
|
36.06 ± 5.40
|
0.486
|
GLOB (g/L)
|
32.47 ± 5.18
|
31.83 ± 7.57
|
0.21
|
30.88 ± 3.85
|
29.40 ± 5.85
|
0.207
|
ALT (U/L)
|
35.24 ± 44.74
|
22.77 ± 17.07
|
0.009
|
33.47 ± 23.72
|
21.24 ± 14.14
|
0.018
|
AST (U/L)
|
45.96 ± 42.45
|
30.16 ± 16.94
|
<0.001
|
64.53 ± 75.75
|
23.83 ± 11.16
|
0.002
|
Scr (μmol/L)
|
56.98 ± 16.87
|
77.15 ± 50.75
|
0.165
|
55.77 ± 15.70
|
72.92 ± 26.50
|
0.396
|
CK (U/L)
|
81.40 ± 40.59
|
83.27 ± 105.39
|
0.234
|
80.20 ± 48.92
|
72.48 ± 66.81
|
0.282
|
CK-MB (U/L)
|
17.22 ± 8.10
|
11.15 ± 4.97
|
<0.001
|
16.65 ± 8.93
|
11.28 ± 7.47
|
0.166
|
LDH (U/L)
|
317.20 ± 104.29
|
207.04 ± 56.64
|
<0.001
|
356.60 ± 214.84
|
185.10 ± 36.97
|
0.001
|
β2-MG (mg/L)
|
2.95 ± 1.17
|
3.38 ± 2.47
|
0.137
|
2.85 ± 1.04
|
2.34 ± 0.74
|
0.263
|
Immunological parameters
|
|
|
|
|
|
|
C3 (g/L)
|
0.89 ± 0.19
|
0.95 ± 0.22
|
0.489
|
0.95 ± 0.21
|
1.13 ± 0.27
|
0.545
|
C4 (g/L)
|
0.22 ± 0.07
|
0.21 ± 0.08
|
0.771
|
0.26 ± 0.07
|
0.25 ± 0.07
|
0.949
|
IgG (g/L)
|
16.14 ± 3.16
|
16.72 ± 7.67
|
0.071
|
13.57 ± 2.94
|
13.33 ± 3.69
|
0.466
|
IgA (g/L)
|
3.72 ± 1.36
|
3.30 ± 1.47
|
0.692
|
2.95 ± 1.25
|
2.81 ± 1.18
|
0.816
|
IgM (g/L)
|
1.66 ± 0.86
|
1.41 ± 0.77
|
0.771
|
1.32 ± 0.79
|
1.32 ± 0.68
|
0.650
|
ANA (1≥80), n (%)
|
20 (80.00%)
|
56 (60.87%)
|
0.075
|
9 (60.00%)
|
5 (17.24%)
|
0.007
|
Arterial blood gas
|
n=20
|
n=59
|
|
n=13
|
n=21
|
|
PH
|
7.44 ± 0.03
|
7.42 ± 0.04
|
0.279
|
7.42 ± 0.06
|
7.43 ± 0.03
|
0.108
|
PaO2 (mmHg)
|
88.75 ± 20.80
|
80.24 ± 16.07
|
0.128
|
85.15 ± 26.87
|
75.19 ± 16.68
|
0.215
|
OI (mmHg)
|
400.48 ± 84.59
|
372.28 ± 81.85
|
0.735
|
365.51 ± 70.14
|
328.25 ± 93.01
|
0.533
|
Lactate (mmol/L)
|
1.59 ± 0.65
|
1.74 ± 0.86
|
0.531
|
1.82 ± 0.62
|
1.74 ± 0.81
|
0.293
|
Pulmonary function tests
|
n=17
|
n=52
|
|
n=10
|
n=9
|
|
FEV1, %predicted
|
80.41 ± 14.63
|
88.56 ± 18.38
|
0.472
|
74.47 ± 23.53
|
85.27 ± 27.57
|
0.628
|
FVC, %predicted
|
79.98 ± 15.25
|
83.07 ± 16.01
|
0.871
|
76.18 ± 21.43
|
78.01 ± 21.46
|
0.849
|
DLco, %predicted
|
56.06 ± 11.32
|
64.21 ± 20.10
|
0.041
|
55.18 ± 16.54
|
71.28 ± 9.70
|
0.122
|
Data are presented as the mean (M)±standard deviation (SD), and p < 0.05 was considered significant.
ADM: amyopathic dermatomyositis; NSIP: nonspecific interstitial pneumonia; OP: organized pneumonia; M: male; F: female; WBC: white blood cell; hsCRP: high sensitivity C-reactive protein; ESR: erythrocyte sedimentation rate; ALB: albumin; GLOB: globin; ALT: alanine aminotransferase; AST: aspartate aminotransferase; Scr: creatine; CK: creatine kinase; CK-MB: creatine kinase and its MB isoenzyme; LDH: lactate dehydrogenase; β2-MG : beta2-microglobin; C3: complement 3; C4: complement 4; IgG: immunoglobulin G; IgA: immunoglobulin A; IgM: immunoglobulin M; ANA: antinuclear antibodies; PaO2: arterial oxygen partial pressure; OI: oxygenation index; FEV1: forced expiratory volume in 1 second; FVC: forced vital capacity; DLco: diffusing capacity of lung for carbon monoxide.
Relationship between ADM-ILD and LDH
The mean serum LDH level was 327.35 ± 155.73 U/L in the ADM-ILD group (n=37), and 201.78 ± 53.29 U/L in the other ILD group (n=121). LDH levels were significantly higher in the ADM-ILD group than in the other ILD groups (p<0.001) (Table 2). To assess the diagnostic value of serum LDH for ADM-ILD, ROC curve analysis was performed and yielded a cut-off level of 235.5 U/L. At this value, the serum LDH level had a sensitivity of 78.38% and specificity of 82.64% for the possibility of ADM-ILD (area under the curve (AUC)=0.854, p<0.001) (Figure 2).
Table 2 Comparison of lactate dehydrogenase between AMD-ILD and other-ILDs.
|
|
LDH level (U/L)
|
|
|
|
|
|
N
|
Mean ± SD
|
F
|
p value
|
|
|
Other-ILDs
|
121
|
201.78 ± 53.29
|
23
|
<0.001
|
|
|
ADM-ILD
|
37
|
327.35 ± 155.73
|
|
|
|
|
Independent samples t-test.
Survival in the NSIP/OP cohort
During the observation period, 8 of the 37 patients (21.62%) with ADM-ILD died. Nine of 92 (9.28%) patients with other clinical subtypes of NSIP died, and one of 29 (3.45%) patients with other OP died during the follow-up. The 1-, 3- and 5-year overall survival rates were 96.00%, 85.26%, and 73.08% for ADM-NSIP patients compared with 97.81%, 92.20%, and 86.35% for control-NSIP patients, respectively. As shown in Figure 3, there was no significant difference in the Kaplan‒Meier survival curves between the groups (log-rank test, p=0.570). For the OP cohort, the 1-, 3- and 5-year overall survival rates were 86.67%, 73.33%, and 73.33% for ADM-OP patients compared with 100%, 100%, and 94.12% for control-OP patients, respectively. Survival curves between the OP groups were significantly different (log-rank test, p=0.007).
Clinical characteristics and prognostic factors of ADM-ILD patients
The demographic, cutaneous manifestation, radiologic pattern, and myositis autoantibody data in 37 ADM-ILD patients are summarized in Table 3. The mean age at ADM-ILD diagnosis was 58.51 ± 15.10 years, and more cases of ADM-ILD occurred among women. Not all patients in our cohort with confirmed ADM reported rash. The most common rash observed in patients was Gottron’ sign (21.62%), followed by cutaneous heterochromia (18.92%). The most common ADM-ILD radiographic pattern, NSIP, was noted in 25 patients (67.57%), followed by the UIP pattern in 15 patients (40.54%). Myositis autoantibodies detected in ADM-ILD patients included myositis-specific antibodies (MSAs) and myositis-associated antibodies (MAAs). The most frequent antibody was anti-MDA5 (37.84%), and the least frequent was anti-Mi-2 (2.70%) in the MSAs. Anti-Ro-52 had the highest positivity rate (45.95%) among the MAAs. In addition, there was also a considerable number of patients with MSAs + MAAs (45.95%) in our cohort. The median follow-up time of the 37 ADM-ILD patients was 37 (IQR 29.5-42.5) months, and 8 patients died from all causes during the observation period. Three patients had an acute-type disease that followed a fulminant course with rapid progression of respiratory failure, and they died within 1 to 2 months. To further explore the factors that were associated with survival in ADM-ILD patients, we conducted univariate and multivariate analyses using the Cox proportional hazards model (Table 4). The continuous parameters (ferritin and LDH) were converted into classification variables. A P value < 0.05 in the univariate analysis was included in the multivariate Cox analysis. As Table 4 shows, the results of the multivariate analysis indicate that skin ulceration (HR 6.85, 95% CI 1.22-38.26, p=0.028) and OI<300 mmHg (HR 6.07, 95% CI 1.20-30.72, p=0.029) were significantly associated with poor prognosis. Despite the substantial number of studies on ADM-ILD reporting that LDH and ferritin correlated with outcome, we failed to confirm the association between the two indicators with increased mortality risk due to the low number of death cases. We found that a ferritin value of less than 500 ng/ml was a protective prognostic factor for ADM-ILD, and the overall mortality rate decreased by 85% (HR 0.15, 95% CI 0.03-0.77, p=0.023) in the univariate analysis. The presence of anti-MDA5 with MAAs appeared to delay disease progression, although the correlation was not significant (p=0.216).
Table 3 Baseline clinical characteristics of the ADM-ILD patients
characteristics
|
ALL patients (n=37)
|
Age
|
58.51 ± 15.10
|
Gender (M/F)
|
14/23
|
Ever smokers, n
|
7 (18.92%)
|
Heliotrope rash, n
|
6 (16.22%)
|
Gottron's sign, n
|
8 (21.62%)
|
Skin ulcer, n
|
4 (10.81%)
|
Mechanic hands, n
|
3 (8.11%)
|
Cutaneous heterochromia, n
|
7 (18.92%)
|
Radiologic pattern
|
|
NSIP
|
25 (67.57%)
|
OP
|
15 (40.54%)
|
NSIP + OP overlap
|
3 (8.11%)
|
Myositis autoantibodies
|
|
MSAs
|
|
Anti-MDA5
|
14 (37.84%)
|
Anti-Jo-1
|
6 (16.22%)
|
Anti-PL-7
|
7 (18.92%)
|
Anti-PL-12
|
3 (8.11%)
|
Anti-EJ
|
4 (10.81%)
|
Anti-SRP
|
2 (5.41%)
|
Anti-Mi-2
|
1 (2.70%)
|
MAAs
|
|
Anti-RO-52
|
17 (45.95%)
|
Anti-Ku
|
2 (5.41%)
|
Anti-PM-Scl 100
|
1 (2.70%)
|
Anti-PM-Scl 75
|
1 (2.70%)
|
Anti-SS-A
|
1 (2.70%)
|
Anti-U1-RNP
|
1 (2.70%)
|
MSAs + MAAs
|
17 (45.95%)
|
ADM: amyopathic dermatomyositis; ILD: interstitial lung disease; M: male; F: female; NSIP: nonspecific interstitial pneumonia; OP: organized pneumonia; MSA: myositis-specific autoantibody; MAA: myositis-associated autoantibody; MDA5: melanoma differentiation-associated gene 5; Jo-1: histidyl-tRNA synthetase; PL-7: threonyl-tRNA synthetase; PL-12: alanyl-tRNA synthetase; EJ: glycyl-transfer ribonucleic acid synthetase; SRP: signal recognition particle; PM-Scl: polymyositis/Scl; MSAs+MAAs: both MSA and MAA positive.
Table 4 Univariate and multivariate analyses of prognostic factors for mortality in ADM-ILD patients.
Variables
|
Univariate
|
|
Multivariate
|
|
|
HR (95%CI)
|
P value
|
HR (95%CI)
|
P value
|
Age
|
1.06 (1.00 – 1.12)
|
0.064
|
|
|
Male
|
2.70 (0.62 – 11.69)
|
0.184
|
|
|
Ever smokers
|
3.53 (0.83 – 15.04)
|
0.088
|
|
|
Skin ulcerations
|
8.93 (1.94 – 41.15)
|
0.005
|
6.85 (1.22 – 38.26)
|
0.028
|
ANA positive
|
0.82 (0.16 – 4.11)
|
0.805
|
|
|
Anti-MDA5 positive
|
1.04 (0.25 – 4.41)
|
0.953
|
|
|
Anti-MDA5 with MAAs positive
|
0.36 (0.07 – 1.81)
|
0.216
|
|
|
OP pattern
|
2.06 (0.50 – 8.56)
|
0.319
|
|
|
OI < 300 mmHg
|
5.15 (1.20 – 22.01)
|
0.027
|
6.07 (1.20 – 30.72)
|
0.029
|
Ferritin < 500 ng/ml
|
0.15 (0.03 – 0.77)
|
0.023
|
--
|
NS
|
LDH ≥ 355 U/L
|
1.21 (0.24 – 6.07)
|
0.812
|
|
|
CI: confidence interval; ANA: antinuclear antibodies; MDA5: melanoma differentiation-associated gene 5; MAA: myositis-associated autoantibody; OP: organized pneumonia; OI: oxygenation index; LDH: lactate dehydrogenase.