Of 31 IPAF subjects enrolled, all of them are on medication in this cohort (Figure 1A), 16 (51.6%) had a history of medicine, 13 (41.9%) had comorbidities, 7 (22.6%) had a history of smoking, and 14 (45.2%) in Stability, 9 (29.0%) in Improvement and 14 (25.8%) in an Exacerbation (Figure 1A). Of 5 IPF subjects, all of them also undergo medication resembling the IPAF cohort, 4 (80.0%) had a history of medicine, 2 (40.0%) had comorbidities, and 1 (20.0%) had a history of smoking. On the strength of this assessment, 1 (20.0%) in a Stability, 2 (40.0%) in an Improvement and 2 (40.0%) in an Exacerbation (Figure 1B). Baseline characteristics of IPAF and IPF cohorts are summarized in Table 1. In comparing those amid the IPAF and IPF cohorts, IPAF patients were inclined to be on medication (Methylprednisolone, Acetylcysteine, and Pirfenidone). In contrast, there was no evident statistical significance in the IPF cohort.
Clinical risk factors
Based upon the T-Test, we found older age, female sex, BMI（> 24）, and even ever-smoker associated with IPAF and IPF (Table 1), among which, seeming that when received medical therapy, the clinical performance of females is better than that of male in stratification of the degree of disease of the severity of IPAF (see Table 2, Figure 2).
In the IPAF cohort, subjects tended to use methylprednisolone (83.9%), followed by acetylcysteine and pirfenidone, whereas in IPF patients seemed to preferentially use acetylcysteine (80.0%) (Table 1). We also found that baseline characteristics of IPAF Subjects Stratified by Severity in Table 2 were all preferred Methylprednisolone. However, in terms of the result of patients of representation, the drug Methylprednisolone is not effective enough. Interestingly, Acetylcysteine was used more frequently in patients with Stability - IPAF and Improvement – IPAF (Table 2).
Matrix Metalloproteinases (MMPs)
Matrix metalloproteinases are a huge family of 23 endogenous zinc-containing proteases, including collagenase, gelatinase, matrix lysozyme, matrix elastase, and membrane matrix metalloproteinases.[34, 35] In humans, this biomarker has a complex relationship with various disease processes, including atherosclerosis, hepatic fibrosis, and interstitial lung fibrosis. In accordance with the result of IPAF cohort, compared with Stability – IPAF, levels of MMP-2, MMP-3, MMP-7, MMP-9, MMP-10, MMP-13 were significantly varied between Improvement - IPAF and Exacerbation – IPAF while the corresponding biomarker to them may be found in this study. In addition, we observed a negative correlation between Levels of MMP-12 and the progression of Improvement - IPAF and Exacerbation – IPAF when compared with Stability - IPAF in IPAF specimens (Table 2). The AUCs of MMP-2, MMP-3, MMP-7, MMP-9 and MMP-13 for Stability AUC vs. Improvement AUC and Stability AUC vs. Exacerbation AUC were 0.683, 0.587, 0.512, 0.643, 0.603 and 0.491, 0.625, 0.634, 0.580, 0.741, respectively (Table 3). When using MMPs as a total factor, the AUC increased to 0.619 and 0.643 for Stability AUC vs. Improvement AUC and Stability AUC vs. Exacerbation AUC in the IPAF cohort (P<0.05 for the difference between the curves). Moreover, we found that MMP-2 and MMP-9 had a better utility in Stability AUC - Improvement AUC than Stability AUC - Exacerbation AUC, nevertheless, MMP-3, MMP-7, and MMP-13 obtained a contrary outcome between the above respective comparison (see Figure 3). When adding MMPs to combinatorial biomarkers (SP-A, KL-6, LDH, CRP, tIgE), the ROC curve AUCs of Stability AUC - Improvement AUC followed similarly, but Stability AUC - Exacerbation AUC showed a stronger identification trend. Interestingly, the AUC of the MMP-12 variable ranged from 0.730 to 0.737 in Stability AUC vs. Improvement AUC and Stability AUC vs. Exacerbation AUC, with the strongest degree of disease progression correlation (see Table 3).
Notably, in the IPF cohort, a great deal of previous research into MMPs has focused on levels of MMP-3, MMP-7, MMP-8, MMP-9, MMP-12, and MMP-13.[37, 38] Owing to the fact that the number of subjects in this cohort is rare, more experiments are needed to be proved in future studies.
Other investigational biomarkers
Levels of KL-6 and SP-A significantly increased with severity of IPAF, among which, KL-6 levels peaked in the IPAF cohort. Meanwhile, CRP and tIgE may be significantly stronger associated with IPF based on the T-test (Table 1). However, due to a small number of patients in the IPF cohort, this conclusion remains to be verified. Multivariable logistic regression analyses adjusting for the above five investigational biomarkers in IPAF Subjects stratified by severity of IPAF are presented in Table 2. In addition to KL-6, SP-A increased significantly, and the other three investigational biomarkers showed no obvious abnormalities (all within the normal range).
Given the degree of disease stratification of IPAF, AUCs for the five investigational biomarkers ranged from 0.379 to 0.710, and these combinatorial biomarkers were 0.583 and 0.867, respectively (Table 3). In Stability AUC vs. Improvement AUC, the AUCs ranged from 0.537 to 0.710 with a combined AUC of 0.583, in Stability AUC vs. Exacerbation AUC, the AUCs ranged from 0.379 to 0.600 with a combined AUC of 0.867 (see Figure 4).
A combination of clinical risk factors and MMPs is strongly associated with IPAF, including Stability, Improvement, and Exacerbation. Importantly the addition of the five investigational biomarkers SP-A, KL-6, LDH, CRP, and tIgE significantly increased the AUC to 0.756 for comparison of Stability AUC - Exacerbation AUC in IPAF cohort (P < 0.05 for the difference between the curves) (Table3, Figure 4). [27, 39-49]
Pulmonary function testing (PFT) and Blood cell test
Whether IPAF is compared with IPF, or simply a comparison among the three stratifications of the severity of IPAF (p<0.05), apart from DLCO, which identified 36 patients with moderate interstitial lung disease, the remaining PFT items, including FVC, FEV1 showed that the patients in two cohorts did not show significant clinical significance in PFT (Figure 5). No abnormality was found in the selected cytokines in the blood cell test in the IPAF and IPF cohorts.