Baseline characteristics
In total, 337 patients with IIPs were screened, and 54 were excluded because of missing data (n = 41) and AE at the time of diagnosis (n = 13), leaving 283 patients (Fig. 1). The clinical characteristics of these patients are presented in Table 1. Sixty-eight (24.0%) patients underwent surgical lung biopsy. Additionally, 94 (33.2%), 19 (6.7%), 16 (5.6%), 12 (4.2%) and 4 (1.4%) patients were diagnosed with IPF (clinical, n = 66; pathological, n = 28), COP, NSIP, pleuroparenchymal fibroelastosis (PPFE) and desquamative interstitial pneumonia (DIP)/respiratory bronchiolitis-associated interstitial lung disease (RB-ILD), respectively (Table 2). The remaining 138 patients (48.7%) had unclassifiable IIPs. Thirty-nine (13.7%) patients had autoimmune features, including 6 (17.4%), 39 (100%) and 36 (92.3%) patients in the clinical, serologic and morphologic domains, respectively. The median observation time was 39.4 months (range, 1.0–165.0 months).
Table 1
Characteristics of the study patients
| All, n = 283 | Cluster I, n = 35 | Cluster II, n = 62 | Cluster III, n = 89 | Cluster IV, n = 97 |
Age | 70.6 (20–90) | 70.7 (52–83)# | 75.9 (55–90)$, ¶ | 68.9 (20–89) | 69.2 (41–88) |
Sex, male | 214 (75.6) | 22 (62.8)$ | 49 (79.0)$, ¶ | 83 (93.2)¶ | 60 (61.8) |
Body mass index | 22.8 (13.1–36.8) | 22.9 (18.1–29.7)# | 20.8 (13.1–25.8)$, ¶ | 24.9 (19.4–36.8)¶ | 22.1 (14.9–29.8) |
Smoking history | 182 (64.3) | 19 (54.2)$ | 31 (50.0)$ | 79 (88.7)¶ | 53 (54.6) |
Pack-year smoking | 42 (2.5–165) | 30 (18.8–66)$ | 40 (5-100)$ | 49 (5-165)¶ | 40 (2.5–90) |
Dust exposure Organic materials Inorganic material | 69 (24.3) 2 (0.7) 67 (23.6) | 10 (28.5)# 0 (0) 10 (28.5) | 6 (9.6)$ 0 (0) 6 (9.6) | 40 (44.9)¶ 1 (1.1) 39 (43.8) | 13 (13.4) 1 (1.0) 12 (12.3) |
Autoimmune features | 39 (13.7) | 30 (85.7) | 6 (9.6) | 1 (1.1) | 2 (2.0) |
Spirometry | | | | | |
% predicted FVC | 78.5 (27.6–124.0) | 81.8 (31.6–113.0)# | 64.5 (27.6–116.0)$, ¶ | 73.0 (39.8–112.0)¶ | 89.1 (40.9–124.0) |
% predicted FEV₁ | 81.0 (31.9-131.6) | 82.3 (66.1–114.0)#, $ | 68.9 (31.9–100.0)$, ¶ | 72.4 (41.4–117.0)¶ | 87.6 (61.0-131.6) |
FEV₁/FVC | 82.4 (53.4–100.0) | 81.7 (65.8–100.0) | 87.5 (63.4–100.0) $ | 81.0 (53.4–98.7) | 82.3 (61.3–98.8) |
Laboratory data | | | | | |
CRP (mg/dL) | 0.20 (0.01–25.5) | 0.27 (0.01–4.8)#, ¶ | 1.13 (0.02–25.5)$, ¶ | 0.23 (0.02-9.0)¶ | 0.10 (0.01–4.7) |
LDH (U/L) | 225 (94–569) | 230 (94–319)¶ | 213 (132–450)$ | 241 (170–569)¶ | 205 (141–334) |
Albumin (g/dL) | 4.1 (1.9–4.9) | 4.0 (2.4–4.9)#, $, ¶ | 3.6 (1.9–4.5)$, ¶ | 4.1(2.8–4.8)¶ | 4.2 (3.4–4.9) |
KL-6 (U/mL) | 829 (112–5710) | 1010 (164–3651) | 664 (112–2070)$ | 1148 (177–5710)¶ | 644 (196–3378) |
SP-D (ng/mL) | 179 (17.2–1160) | 166 (25.4–533)$ | 173 (33.4–1130)$ | 218 (49.8–1160)¶ | 154 (17.2–525) |
CT findings | | | | | |
Emphysema | 90 (31.8) | 11 (31.4) | 14 (22.5)$ | 35 (39.3) | 30 (30.9) |
Honeycombing | 82 (28.9) | 5 (14.2)#, $ | 23 (37.0)¶ | 34 (38.2)¶ | 20 (20.6) |
Treatments | | | | | |
Steroids | 96 (33.9) | 15 (42.8) | 21 (33.8) | 36 (40.4)¶ | 24 (24.7) |
Immunosuppressants | 35 (12.3) | 4 (11.4) | 5 (8.0) | 17 (19.1) | 9 (9.2) |
Antifibrotic agents | 56 (19.7) | 3 (8.5)$ | 13 (20.9) | 28 (31.4)¶ | 12 (12.3) |
Data are presented as the median (range) or number (%). |
CRP, C-reactive protein; FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; IIPs, idiopathic interstitial pneumonias; IPAF, interstitial pneumonia with autoimmune features; KL-6, Krebs von den Lungen-6; LDH, lactate dehydrogenase; SP-D, pulmonary surfactant protein-D |
#p < 0.05 compared with Cluster II |
$p < 0.05 compared with Cluster III |
¶p < 0.05 compared with Cluster IV |
Table 2
Characteristics of the study patients
| All, n = 283 | Cluster I, n = 35 | Cluster II, n = 62 | Cluster III, n = 89 | Cluster IV, n = 97 |
IPF | 94 (33.2) | 6 (17.1) | 22 (35.4) | 46 (51.6) | 20 (20.6) |
NSIP | 16 (5.6) | 5 (14.2) | 0 (0) | 5 (5.6) | 6 (6.1) |
COP | 19 (6.7) | 3 (8.5) | 8 (12.9) | 3 (3.3) | 5 (5.1) |
DIP / RB-ILD | 4 (1.4) | 2 (5.7) | 0 (0) | 0 (0) | 2 (2.0) |
PPFE | 12 (4.2) | 0 (0) | 8 (12.9) | 0 (0) | 4 (4.1) |
Unclassifiable IIPs | 138 (48.7) | 19 (54.2) | 24 (38.7) | 35 (39.3) | 60 (61.8) |
Data are expressed as number (%). |
COP, cryptogenic organizing pneumonia; DIP, desquamative interstitial pneumonia; IIP, idiopathic interstitial pneumonia; IPF, idiopathic pulmonary fibrosis; NSIP, non-specific interstitial pneumonia; PPFE, pleuroparenchymal fibroelastosis; RB-ILD, respiratory bronchiolitis-associated interstitial lung disease |
Clinical characteristics of the clusters
Four clusters were identified in the cluster analysis using clinical data (Supplementary Fig. 1).
Cluster I (n = 35) was an intermediate-aged cohort (70.7 years) with low proportions of males (62.8%) and smokers (54.2%; Table 1, Fig. 2A). Cluster I featured preserved %FVC (81.8%), the lowest honeycombing rate (14.2%) and the highest rate of autoimmune features (85.7%). Regarding laboratory data, Cluster I had the second highest KL-6 level (1010 U/mL), a slightly increased SP-D level (166 ng/mL) and normal CRP, LDH and albumin levels. In Cluster I, six (17.1%), five (14.2%), three (8.5%) and two (5.7%) patients were diagnosed with IPF (pathological, n = 3; clinical, n = 3), NSIP, COP and DIP, respectively. The remaining 19 (54.2%) patients had unclassifiable IIP (Table 2).
Cluster II (n = 62) had the oldest population (75.9 years), the smallest percentage of smokers (50.0%), the lowest BMI (20.8) and the lowest %FVC (64.5%; Table 1, Fig. 2B), as well as the second largest male population (79.0%). Emphysema was least frequent in this group (22.5%), but the second highest rate of honeycombing was noted (37.0%). In laboratory findings, Cluster II had the highest CRP level (1.13 mg/dL), the lowest serum albumin level (3.6 g/dL) and slightly increased KL-6 (664 U/mL) and SP-D levels (173 ng/mL). In total, 22 (35.4%), 8 (12.9%) and 8 (12.9%) patients were diagnosed with IPF (clinical, n = 20; pathological, n = 2), PPFE and COP, respectively. The remaining 24 (38.7%) patients had unclassifiable IIP (Table 2).
Cluster III (n = 89) had the youngest population (68.9 years), the highest proportion of males (93.2%), the highest BMI (24.9) and the highest frequencies of smoking (88.7%) and dust exposure (44.9%; Table 1, Fig. 2C). This cluster had the second lowest %FVC (73.0%). Cluster III had the highest LDH (241 U/L), KL-6 (1148 U/mL) and SP-D levels (218 ng/mL) and normal CRP and albumin levels. The rates of emphysema and honeycombing were highest in this group (39.3 and 38.2%, respectively). In total, 46 (51.6%), 5 (5.6%) and 3 (3.3%) patients were diagnosed with IPF (clinical, n = 29; pathological, n = 17), NSIP and COP, respectively. The remaining 35 (39.3%) patients had unclassifiable IIP (Table 2).
Cluster IV (n = 97) had the second youngest population (69.2 years), the highest proportion of females (38.2%), normal BMI (22.1) and lower frequencies of smoking (54.6%) and dust exposure (13.4%; Table 1, Fig. 2D). Cluster IV had the highest %FVC (89.1%) and normal CRP, LDH and albumin levels. Cluster IV had the lowest KL-6 (644 U/mL) and SP-D levels (154 ng/mL). The rates of emphysema and honeycombing were 30.9 and 20.6%, respectively. Overall, 20 (20.6%), 6 (6.1%), 5 (5.1%), 4 (4.1%) and 2 patients (2.0%) were diagnosed with IPF (clinical, n = 13; pathological, n = 7), NSIP, and COP, PPFE and DIP, respectively. The remaining 60 (61.8%) patients had unclassifiable IIPs (Table 2).
Differences in clinical outcomes among the clusters
OS was significantly longer in Cluster IV than in Clusters II (p < 0.01) and III (p = 0.01; Fig. 3A). Clusters I and III had significantly longer OS than Cluster II (both p < 0.01). The 5-year OS rates were 87.7, 79.0, 70.1 and 32.3% in Clusters IV, I, III and II, respectively, and the 10-year OS rates were 66.6%, 57.6%, 41.4% and 32.3%, respectively.
Clusters II and III had significantly higher cumulative incidences of AE than Cluster IV (both p = 0.03; Fig. 4A). The 5-year cumulative incidence rates of AE were 20.6%, 19.9%, 9.7% and 8.0% in Clusters II, III, I, and IV, respectively, and the 10-year cumulative incidence rates were 20.6%, 23.0%, 21.7% and 13.8%, respectively.
Next, OS and AE were evaluated according to the IIP diagnosis. Patients with NSIP had significantly longer OS than those with IPF (p = 0.01), whereas there was no significant difference among the other groups (Fig. 3B). Patients with IPF had a significantly higher cumulative incidence of AE than those with COP (p = 0.01) and unclassifiable IIPs (p < 0.01), whereas there was no significant difference among the other groups (Fig. 4B).