Development of a nomogram for predicting the presence of combined pulmonary fibrosis and emphysema

Pulmonary interstitial fibrosis and emphysema have long been perceived as two separate diseases. Interstitial lung diseases (ILD) encompass a large and heterogeneous group of diffuse parenchymal lung disorders characterized by distinct forms and severity of inflammation and fibrosis in alveolar walls and cavities [1]. Idiopathic pulmonary fibrosis (IPF), an ILD of unknown cause is invariably progressive and associated with poor prognosis [2]. Emphysema, most often caused by long-term exposure to cigarette smoke, featured with abnormal and permanent enlargement of air spaces distal to the terminal bronchioles, is one of the major pathobiological processes leading to chronic obstructive pulmonary disease [3]. Progressive inflammation damages the airway mucosal epithelium, which in turn leads to airflow limitation and lung parenchymal destruction [4]. Therefore, emphysema and fibrosis are often considered distinct entities with unique pathophysiologic manifestations, but in the past 15 years, there has been an increasing recognition that these two processes may coexist in individual patients. “Combined pulmonary fibrosis and emphysema (CPFE)” was first described as a well-defined syndrome by Cottin et al. in 2005 [5].

Patients with CPFE are characterized by a relatively normal lung function due to the counterbalancing effects of fibrotic (restrictive factor) and emphysematous (obstructive factor) components [6, 7], which often lead to underestimating the severity of CPFE, or even a delayed or missed clinical diagnosis. A previous study conducted by Mejía M et al. reported that in the series of the 110 patients initially diagnosed with IPF, 28% were reevaluated and classified as CPFE [8]. The other study found that CPFE was found in 33.5% of 660 patients with usual interstitial pneumonia (UIP) [9]. Although computed tomography (CT) scan of the chest is routinely performed in patients with IPF, the development of emphysema is considered as a long cumulative process, which gas exchange and mechanical abnormalities may predate radiographic low attenuation areas of the lung parenchyma. Consequently, the presence of CPFE in patients diagnosed with IPF is of concern. In addition, CPFE is frequently complicated by pulmonary hypertension [10], lung cancer [11], acute exacerbations [12], and leading to poor natural history and prognosis [13]. Currently, there is still a lack of specific drugs for clinical treatment [14]. Considering this, a predictive model with reliable efficacy is of great importance to helps us raise the profile of patients with possible CPFE early (e.g., before imaging, or some who refused frequent CT scans), so as to conduct appropriate clinical treatment of CPFE.

The nomogram provides a visualization of the regression equation, which has been accepted as a reliable tool to create a simple intuitive graph of a statistical predictive model that quantifies the risk of a clinical event [15]. In this work, we performed a retrospective study to create an easy-to-use risk assessment nomogram model integrating multiple clinical risk factors for predicting the presence of CPFE to support clinicians in their treatment recommendations.

CPFE is considered as a not fully recognized syndrome characterized by chronic, progressive disease with worsening respiratory symptoms, reduced lung function and poor prognosis [5]. The analysis of risk factors has guiding significance for the early recognition, clinical diagnosis and appropriate treatment. Here, this study describes the clinical characteristics and incorporates multiple clinical variables into a user-friendly nomogram for predicting the presence of CPFE.

With the presence of fibrosis and emphysema concomitantly, pulmonary function tests of CPFE patients are characterized by the preservation of lung volumes and markedly impaired carbon monoxide diffusing capacity, rather than the simple coexistence [21]. The relatively normal lung volumes in CPFE usually result from the counterbalancing effects of the restrictive disorder of fibrosis and the hyperinflation of emphysema [22], meanwhile, the presence of these two factors leading to severe reduction in the amounts of functional alveolar-capillary units [23]. Our results indicated that CPFE patients showed significantly higher lung volume (RV, VC, and TLC) and ventilation indicators (VA, FVC, and FEV1), and DLCO/VA were much more decreased, consistent with previous studies. Moreover, positive correlation was observed between DLCO and VA%, RV%, TLC% in patients with IPF but not in patients with CPFE, which remind us more clinical attention on the variation consistency of lung volume, ventilation and diffusion function indicators.

Most patients with CPFE are males, and they are either current or ex-smokers [24, 25]. Current study demonstrated that smoking has been considered as a risk factor for the development of CPFE [26]. The results of this study came to the same conclusion. The mechanism behind this could also be due to a sequence of events that first cause bronchial inflammation, small airway stenosis and alveolar rupture leading to emphysema; and then additionally stimulate the epithelial mesenchymal transition (EMT) to promote the differentiation of fibroblasts into myofibroblasts [25]. In addition, late-onset increased gastroesophageal reflux (GER) triggered by smoking may also aggravate the fibrotic changes [27].

The higher proportion of patients with allergies in the CPFE group indicated that immune mediators (mast cells, basophils, eosinophils, cytokines, chemokines, etc.) may be associated with the development of disease. Several studies have demonstrated that air contamination is closely related to emphysema and pulmonary fibrosis [28]. Allergy-prone patients who have more abundant and expressed IgE and FcεR receptors [29] may experience more significant inflammation and immune responses when exposed to airborne antigens. Furthermore, TGF-β1, for example, is known to play an major role in the differentiation of fibroblasts into myofibroblasts [30] and eosinophil-derived IL-13 is closely associated with emphysema [31]. It can be speculated that immune impairment in CPFE patients may be more pronounced due to the superimposed effect of immune damage in pulmonary fibrosis and emphysema.

Smoking is undoubtedly a main factor, but not all CPFE patients in this study had a history of smoking (88%). On the one hand, patients with second-hand smoke who are not easily defined may be ignored; on the other hand, it is currently believed that multiple factors are involved in the development of CPFE. Besides inflammation, gene-mediated alveolar damage processes may also lead to CPFE [29]. Oxidative stress and accelerated lung aging with telomere shortening has been proposed as a possible mechanism related to CPFE pathogenesis as well [32]. Another theory suggests that fibrosis occurs predominantly at the base of the lung and can cause local lobe contraction, with progressive compensatory emphysematous changes in the upper lobes of the lung affected by tensile forces [33], consistent with the classical imaging features of CPFE. In short, further studies are needed to shed light on the pathogenesis of CPFE.

In our study, we use multivariable logistic regression analysis to identify significant factors associated with CPFE. Consequently, gender (male), smoking, allergies, FEV1/FVC%, DLCO/VA% pred were identified and used to develop the prognostic nomogram. This nomogram demonstrated good discrimination as assessed by the C-index, AUC value and calibration plot indicating good performance. Nomogram models are used to assess the risks associated with CPFE and they also provide a reference for the clinical management.

Although our study lies in the intuitive characteristics of the disease based on the real-world data and the relatively complete information, which can ensure the accuracy of the model, there are still some limitations. First, the data for the nomogram were retrospectively derived from a single center and may suffer from selection bias. Second, only internal verification was performed and the results may overestimate the effectiveness of the model. Thus, external verification will be optimal for further investigation.

In conclusion, our nomogram incorporating several important clinical variables into the estimate of the risk of CPFE may serve as a potential tool to help inform decision-making by physicians and patients.