In this study, the CTD-ILD group had more Raynaud`s phenomenon which was a risk factor for developing ILD in multivariable logistic regression analysis. Previous studies mentioned different prevalence of Raynaud`s between idiopathic ILD and CTD-ILD [12]. However, it has been sporadically reported Raynaud`s phenomenon as a risk factor for developing ILD in CTD. Xie et al. reported that Raynaud’s phenomenon was a risk factor for developing ILD in IIM [13]. Narula et al. reported that Raynaud`s phenomenon had a positive association with the development of ILD in mixed connective tissue diseas (MCTD) [14].
On radiologic HRCT evaluation in this study, 58.6% of CTD-ILD patients had NSIP pattern and 35.7% had UIP pattern, consistent with previous studies. The two most common histologic patterns are UIP and NSIP. UIP is the predominant pattern in IPF, and NSIP is seen more in CTD-ILD [15–17]. In our study, RA-ILD had a larger proportion of UIP. Up to 60% of RA cases had the UIP pattern [18]. In SSc, IIM, and SLE, the predominant pathologic pattern is NSIP, and overall survival in this cohort is more favorable than for UIP associated with a CTD [19]. When stratified according to CTD subgroup, the proportion of UIP pattern is consistent with previous studies.
There were no significant differences in clinical findings according to CTD subtype except for pulmonary function test in IIM. In our study, IIM-ILD had poorer pulmonary function than other CTD types. There was no well-designed study comparing pulmonary function between each type of CTD-ILD. However, this difference in function could be explained by differences in the clinical courses of IIM-ILD and other CTD-ILD. The clinical course of IIM-ILD can be categorized into three clinical patterns: rapidly progressive, chronic and asymptomatic form [20]. In contrast, most patients with other CTD-ILD, have chronic progression. This difference reflects that inflammatory change is more prominent in myositis-associated ILD and is generally proportional to the rate of symptom deterioration, with greatest severity in the rapidly progressive form [21].
Most cases of CTD-ILD are diagnosed in patients with a rheumatologic diagnosis of a well-established CTD. A substantial minority of patients, present with ILD first, and CTD is diagnosed at a later date [22–27]. In our study, 17.1% of patients were diagnosed with ILD first, consistent with previous research. In a previous study, up to 15% of patients initially diagnosed with idiopathic NSIP had underlying CTD on further investigation [12]. We tried to identify clinical risk factors present by the time ILD occurs. However there were no significant differences in clinical features. For future research, further well-designed prospective studies are warranted to uncover detailed clinical factors affecting the development of ILD in patients who have been diagnosed with CTD first.
CTD-ILD is known to have a better prognosis than idiopathic interstitial pneumonia [28–31]. In SSc, the predominant pathologic pattern is NSIP, and overall survival in this cohort is more favorable than in UIP [21]. However, recent data have shown that CTD-ILD has varying prognoses depending on the type of underlying CTD or ILD pattern. For example, the presence of UIP in association with any CTD portends a prognosis that rivals the mortality associated with IPF [28, 29]. Unlike idiopathic ILD, in which NSIP was associated with better survival than UIP, CTD-NSIP and CTD-UIP have shown similar prognoses [30, 31], except for RA-ILD. Therefore, the exact evaluation of CT pattern is important to predict prognosis and provide proper management. In our study, the ILD-preceding group had poorer lung function at initial diagnosis and more decreased pulmonary function at follow up. As far as we know, there is no previous study on the prognosis of CTD-ILD stratified by ILD onset time. This study has implications for us that it is necessary to looking for clues indicating CTD in patients with already diagnosed ILD.
IPF is characterized by progressive lung remodeling associated with excessive deposition of extracellular matrix (ECM) [32]. The contribution of MMPs to the ECM remodeling is very complex because MMPs have multiple effects. They are not only responsible for ECM degradation, but also cleave diverse bioactive mediators such as growth factors, cytokines, and chemokines to release them from the ECM [33]. MMP-7, the smallest MMP (247 amino acids), plays a crucial role in pulmonary inflammation and fibrosis. MMP-7 is expressed in alveolar macrophages and bronchiolar epithelial cells. MMP-7 levels were shown to be elevated in serum and bronchoalveolar lavage (BAL) fluid from patients with IPF compared with healthy controls [34, 35]. Rosas et al. reported that MMP-7 was significantly higher in IPF populations than in healthy controls. They documented an increased in MMP-7 in lung tissue and BAL fluid obtained from patients with IPF. In addition, they reported that the elevated MMP-7 levels were negatively correlated with percent FVC% predicted and DLCO% predicted [36]. In IPF, plasma MMP-7 levels could predict mortality [37].
The present study demonstrated that MMP-7 was the most significant cytokine in ILD with CTD. MMP-7 levels were associated with the development of ILD in CTD patients in the present study. Also, MMP-7 had a significant association with CT score and a negative correlation with DLCO% in CTD-ILD. The role of MMP-7 in ILD associated with CTD has been reported in some studies. Chen et al. reported similar findings in RA-ILD that serum levels of MMP-7 are significantly elevated in RA-ILD [9]. In a recent study, Nakatsuka et al. reported that MMP-7 levels were higher in anti-amimoacyl-tRNA synthetase (ARS) antibody positive IIM-ILD; they also reported higher serum MMP-7 levels among anti-melanoma differentiation-associated protein 5 (MDA5) antibody positive IIM-ILD patients were associated with a worse prognosis [38]. To our knowledge, this is the first study evaluating the correlation of MMP-7 in the entire CTD-ILD. Although there is no significant association with ILD progression in this study, further prospective studies with larger cohorts are needed on this point.
The results for IL-6, IL-8, IL-10, and IP-10 were different from those of previous studies. In our study, IL-6, IL-10, and IP-10 were significantly higher in the CTD without ILD group. This is thought probably because the underlying disease activity was not reflected. Differences in the levels of these cytokines and chemokine were not significant after adjustment for inflammatory markers, including ESR and CRP, and disease activity score (DAS28-ESR) in RA-ILD. IL-6, a representative pro-inflammatory cytokine, contributes to the pathogenesis of rheumatologic diseases. In RA, it has been demonstrated that IL-6 is involved in local inflammation that causes joint destruction and various systemic inflammatory signs and symptoms [39]. IL-6 is found in the serum of patients with RA and its level correlates with the disease activity [40]. IL-10, a potent cytokine synthesis inhibitory factor and anti-inflammatory cytokine, plays a central role in the pathogenesis of autoimmune diseases [41]. High levels of serum IL-10 detected in RA patients is reported in previous studies [42]. A more significant increase in the level of IL-10 in patients having high disease activity (DAS > 3.2) shows that the level of IL-10 rises in response to higher inflammatory state of these patients [43]. The chemokine IP-10 regulates immune responses by activation and recruitment of leukocytes through its receptor, CXCR3 [44]. Increased serum levels of IP-10 were associated with inflammatory responses in previous studies. Serum IP-10 levels are elevated prior to the development of RA [45]. IP-10 levels also have significant correlation with disease activity scores (DAS) in RA [46].
This is the first study investigating the radiologic features and respiratory function of CTD-ILD in Korea regarding to onset time of ILD. Our findings provide insight that ILD preceding group has more deteriorated lung function, and patients with IIM have more chance to develop ILD earlier than CTD. In this study, serum MMP-7 was significantly higher in the CTD-ILD group significantly and had significant correlation with CT grade and DLCO.
Despite some significance, this study has a few limitations. It is difficult to establish a cause and effect relationship, because the study is retrospective and includes a small number of patients. It is also possible that clinical aspects related to autoimmune features including Raynaud`s phenomenon were underreported in the non-ILD patients, leading to a retrospective bias. The prognostic results did not match CT score and PFT. This is because the PFT was not performed at exactly the same time in patients with CT. In CTD-ILD patients, sophisticated analysis of PFT is needed, because CTD can involve the bronchi, pulmonary vessels, and muscular structures in addition to lung parenchyma [47]. As mentioned previously, CTD is a heterogenous group of inflammatory diseases, and disease activity at the time of enrollment could not be analyzed. The last limitation is that not all non-ILD groups were identified by CT.