Acute exacerbation of progressive pulmonary fibrosis: incidence and outcomes

Abstract

Background

Few data are available on acute exacerbation (AE) in patients with progressive pulmonary fibrosis (PPF) besides idiopathic pulmonary fibrosis (IPF). This study aimed to investigate the AE incidence and outcomes among patients with PPF.

Methods

Clinical data of patients with PPF (n = 133) were retrospectively collected at a single center. PPF was defined based on the criteria used in the INBUILD trial. AE was defined as a worsening of dyspnea typically within 30 days with new bilateral lung infiltration and no evidence of cardiac failure or fluid overload.

Results

Among patients with PPF, the mean age was 60.6 years old, 57.1% were females, and the most common etiology was connective tissue disease-related ILDs (63%). During the follow-up (median: 38.0 months) after PPF diagnosis, 42 patients (31.6%) experienced AE. The 1-, 3-, and 5-year AE incidences were 12.5%, 30.3%, and 38.0%, respectively. Older age, rheumatoid arthritis associated ILD, fibrotic hypersensitivity pneumonitis, and lower lung diffusing capacity for carbon monoxide were AE risk factors. Patients with AE demonstrated worse survival (median survival: 30 months vs. not reached; p < 0.001) after PPF diagnosis than those without. AE was independently associated with mortality in patients with PPF (hazard ratio [HR], 2.194; 95% confidence interval [CI], 1.285—3.747; p = 0.004) in the multivariable Cox analysis, along with older age, lower lung diffusing capacity for carbon monoxide, and the usual interstitial pneumonia-like pattern on high-resolution computed tomography.

Conclusions

Our results suggest AE is not uncommon and significantly impacts on survival in patients with PPF.

Background

Interstitial lung disease (ILD) is a heterogeneous lung disease group with various clinical courses [1, 2]. Idiopathic pulmonary fibrosis (IPF) is a prototype of progressive fibrosing ILD, characterized by worsening dyspnea, decline in lung function and quality of life, as well as early mortality [3, 4]. On the contrary, among non-IPF ILDs, most patients have a stable clinical course and good prognosis. However, some of these patients demonstrate a progressive fibrosing phenotype, called progressive pulmonary fibrosis (PPF), especially among certain conditions, including idiopathic nonspecific interstitial pneumonia (iNSIP), unclassifiable idiopathic interstitial pneumonia (IIP), hypersensitivity pneumonitis (HP), and connective tissue disease-related ILDs (CTD-ILD). PPF may develop with an 18—32% frequency among non-IPF fibrosing ILDs, and its clinical course is similar to IPF [5]. In addition to clinical similarities, PPF shares a common pathobiological mechanism with IPF that involves a fibrotic response to tissue damage [6].

Most patients with IPF demonstrate slow progressive deterioration, but some patients experience rapid worsening episodes called acute exacerbations (AE). The annual AE incidence among IPF cases is known to be approximately 5—9%, with an almost 50—80% in-hospital mortality rate [4, 7, 8]. AEs are noted to occur in non-IPF fibrosing ILD cases as well as in IPF [9,10,11,12]. Suda et al. reported that in a study of 83 patients with CTD-ILD, six (7.2%) experienced an AE during the follow-up period (mean: 6.0 ± 5.6 years) and had a high overall mortality rate (83.3%) [12]. A study by Kang et al. involving 101 patients with HP reported that 18 (17.8%) experienced an AE during the follow-up period (median: 30 months) and had a significantly poorer survival than those who did not [9]. Suzuki et al. recently studied 557 patients with non-IPF fibrosing ILD. They reported that although the AE incidence of PPF cases is about half of IPF cases (3.21 and 8.38 per 100 patient-years in non-IPF ILDs and IPF, respectively) during the follow-up period (median: 3.4 years), the AE prognosis is similar to IPF cases [10]. However, data on the AE incidence and prognosis in patients with PPF are still lacking. Therefore, this study aimed to investigate the AE incidence, risk factors, and clinical outcomes in patients with PPF.

Materials and methods

Study population

The study population consisted of 509 patients with non-IPF fibrosing ILD who were consecutively diagnosed with iNSIP (n = 98; all biopsy confirmed), fibrotic HP (n = 76; all biopsy confirmed), and CTD-ILD (n = 335; biopsy confirmed = 85) between January 2005 and December 2015 at the Asan Medical Center, Seoul, Republic of Korea. According to the PPF criteria suggested in the INBUILD trial, patients with less than 2 years of follow-up after ILD diagnosis were excluded (n = 113), including those who died within 2 years. Among patients followed for 2 years or more (n = 396), patients not meeting the PPF criteria were excluded (n = 263). Therefore, 133 patients were included in the analysis (Fig. 1). All participants in this study were included in the previous study [13].

Fig. 1

Study flow chart. ILD: interstitial lung disease, IPF: idiopathic pulmonary fibrosis, AE: acute exacerbation, PPF: progressive pulmonary fibrosis

All patients with iNSIP and fibrotic HP met the diagnostic criteria of the American Thoracic Society (ATS) project and the ATS/Japanese Respiratory Society/Latin American Thoracic Association clinical practice guidelines, respectively [14, 15]. All patients with CTD-ILD met the American College of Rheumatology classification criteria for rheumatoid arthritis (RA), systemic sclerosis (SSc), and Sjogren syndrome (SJS) [16,17,18,19]. All diagnoses were made through multidisciplinary discussions. The study protocol was approved by the Asan Medical Center Institutional Review Board (IRB No.: 2022 − 0116). Informed consent requirements were waived due to the retrospective nature of the study.

Data collection

For all patients, survival and clinical data were retrospectively collected from the medical records and/or National Health Insurance of Korea records at the time of PPF diagnosis (2 years after ILD diagnosis). Forced vital capacity (FVC), total lung capacity (TLC), and the diffusing capacity of the lung for carbon monoxide (DL_CO) were measured according to the ATS/European Respiratory Society recommendations [20,21,22]. These findings were expressed as percentages of the normal predicted values. The 6-minute-walk test (6MWT) was performed according to the ERS/ATS recommendations [23].

Records of follow-up visits, usually occurring every 3–6 months, and hospitalizations were reviewed to identify complication developments, including AE pneumonia, pneumothorax, pulmonary embolism, as well as pulmonary hypertension. Rapid deterioration (RD) was defined as acute worsening of dyspnea, requiring hospitalization with new radiologic abnormalities. Based on the suggested criteria by Collard et al. (2016), an AE was defined as a worsening of dyspnea typically within 30 days with new bilateral pulmonary opacities and no evidence of cardiac failure or fluid overload [7]. AE was further categorized into either idiopathic or triggered AE, depending on whether underlying triggers could be identified. The non-AE RD events included pneumonia, pulmonary hypertension, or pneumothorax. Pneumonia was defined as focal or unilateral lung infiltrations with an identified causative organism. However, in cases where a specific pathogen could not be identified, but there was a strong clinical suspicion of infection clinically (symptoms such as purulent sputum, as well as a rapid and significant response to antibiotics alone), it was also categorized as pneumonia [24]. Pulmonary hypertension (intermediate to high probability) was defined as a ≥ 2.9 m/s maximal tricuspid regurgitation velocity according to the 2022 European Society of Cardiology/ERS guidelines [25].

Definition of PPF

PPF was defined based on the INBUILD criteria [26], as a fibrosis extent of more than 10% of the total lung on HRCT, with one of the following criteria within 24 months after diagnosis despite standard of care: (i) a relative decline of ≥ 10% FVC; (ii) a relative decline of 5–10% FVC and respiratory symptom worsening or increased fibrosis extent on HRCT; or (iii) respiratory symptoms worsening and increased fibrosis extent on HRCT. Respiratory symptom worsening was defined as an increase of one or more points in the modified Medical Research Council dyspnea scale.

Radiologic assessment

High-resolution computed tomography (HRCT) scans were obtained according to the standard protocols at full inspiration without contrast enhancement. HRCT images of all patients were evaluated by a thoracic radiologist (J.C.) blinded to the clinical information [13]. HRCT findings classified HRCT patterns as usual interstitial pneumonia (UIP)-like or non-UIP-like patterns. A UIP-like pattern was defined as reticular abnormalities and traction bronchiectasis with or without honeycombing, a basal as well as peripheral predominance, and the absence of atypical features, including extensive ground-glass opacity, nodules, or consolidation [26].

Statistical analysis

All data are expressed as mean ± standard deviation or median (interquartile ranges) for continuous variables and percentages for categorical variables. The Student’s t-test was used for continuous data, and Pearson’s chi-square or Fisher’s exact tests were used for categorical data. The Kaplan–Meier estimates, and the log-rank test were used for survival analysis and cumulative AE incidence rates. The follow-up time was calculated from the PPF diagnosis date to the date of death or censoring time (vital status ascertainment date: 30 June 2019). A Cox proportional hazard analysis analyzed risk factors for AE or all-cause mortality. In the unadjusted analysis, variables with a p < 0.1 were entered into multivariable models. Logistic regression analysis was used to determine risk factors for AE and in-hospital mortality. P-values < 0.05 were considered statistically significant. Statistical analysis was performed using the R software version 4.2.1 (the R Foundation, Vienna, Austria).

Results

Incidence and baseline characteristics

Of all the patients, the mean age was 60.6 years, and 57.1% were female (Table 1). The most frequently observed diagnosis among all patients was CTD-ILD (63.9%), followed by fibrotic HP (21.1%) and iNSIP (15.0%). Among CTD-ILD cases, RA-ILD was the most common (67.1%), followed by SSc-ILD (21.2%) and SJS-ILD (11.8%). None of the patients received antifibrotic treatment.

Table 1 Comparison of baseline characteristics between the RD and non-RD groups among patients with PPF

The median follow-up duration was 38.0 months (interquartile range [IQR]: 20.5–63.0 months) after the PPF diagnosis. During the follow-up period, 58 patients (43.6%) experienced RD. Among these cases, AEs were most common (42 patients [31.6% of all patients]); of these, 17 were idiopathic (12.8%), and 25 were triggered (18.8%). Other occurrences included pneumonia (7.5%), pneumothorax (3.0%), and pulmonary hypertension (1.5%). The 1- and 5-year cumulative RD incidence rates were 16.4% and 49.2%, respectively. Furthermore, the 1- and 5-year cumulative AE incidence rates were 12.5% and 38.0%, respectively (Fig. 2). AEs occurred more frequently in winter or spring (65%, summer or fall: 44%).

Fig. 2

Cumulative incidence of RD and AE in patients with PPF. RD: rapid deterioration, AE: acute exacerbation, PPF: progressive pulmonary fibrosis

AE risk factors

At baseline, the RD group exhibited significantly lower FVC as well as DLco, a shorter 6-minute walk distance (6MWD), lower minimum oxygen saturation (SpO₂) during the 6MWT, and more frequent RA-ILD than the non-RD group (Table 1). Additionally, patients with AE presented comparable characteristics with the RD group. However, regarding FVC and the patient proportion with RA-ILD, those with an AE demonstrated similar trend to the RD group, but there were no statistically significant differences when compared to the non-RD group.

In the unadjusted Cox regression analysis, older age, lower FVC as well as DLco, shorter 6MWD, lower minimum SpO₂ during the 6MWT, and RA-ILD were associated with AE development (Table 2). In the multivariable Cox regression analysis, older age, RA-ILD, fibrotic HP, and lower DLco were independent risk factors for AE (Table 2).

Table 2 Risk factors for AE in patients with PPF

Table 3 Prognostic factors for overall mortality in patients with PPF

Effect on overall survival

During the follow-up after PPF diagnosis, 58 (43.6%) patients died (the median survival: 77.0 months, 95% CI: 52.7—101.3 months). After the PPF diagnosis, patients who experienced an AE demonstrated worse survival with a median survival of 30.0 months, compared to the non-RD group (not reached, p < 0.001) (Fig. 3A). The 1-, 3-, and 5-year mortality rates in the AE group were 21.4%, 65.6%, and 86.2%. These findings were significantly higher than those in the non-RD group (12.0%, 19.8%, and 32.7%, respectively).

Fig. 3

Overall survival curves in patients with PPF. (A) Comparison of survival curves after PPF diagnosis according to clinical course in patients with PPF (B) Comparison of survival curves after hospitalization between triggered AE and idiopathic AE groups among patients with PPF. AE: acute exacerbation, PPF: progressive pulmonary fibrosis, RD: rapid deterioration

In the non-AE RD group, the median survival was 38.0 months, which was significantly shorter than that of the non-RD group (not reached, p < 0.001), and comparable to the survival observed in the AE group (Fig. 3A). The 1-, 3-, 5- year mortality rates were 12.9%, 62.3%, 75.0% respectively, significantly higher than those in the non-RD group.

In the unadjusted Cox analysis, AE was significantly associated with mortality in patients with PPF, along with older age, lower lung function (FVC, DLco), shorter distance, lower minimum SpO₂ during the 6MWT, RA-ILD, and a UIP-like pattern (Table 3). In the multivariable analysis, AE was independently associated with mortality (hazard ratio, 2.194; 95% CI, 1.285–3.747; p = 0.004), along with older age, lower DLco, and a UIP-like pattern on HRCT (Table 3).

Survival after hospitalization

Of the 42 patients with an AE requiring hospital admission, 15 (35.7%) died during hospitalization. Furthermore, the 30-and 90-day mortality rates after the AE were 35.7% and 41.4%, respectively. During hospitalization, the non-survivors (in-hospital) exhibited older ages, lesser frequent male, lower oxygen partial pressure to the fraction of inspired oxygen (PF) ratios, and higher LDH levels than survivors (Table 4). The non-survivors received steroid pulse therapy more frequently than survivors; however, the two groups did not differ in terms of high and low-dose steroid therapy (e-Table 1). In the unadjusted logistic regression analysis, older age, lower PF ratio, and higher LDH levels were significantly associated with in-hospital mortality (Table 5). In the multivariable analysis, lower PF ratios (odds ratio, 0.990; 95% CI, 0.982—0.998; p = 0.015) and older age (odds ratio, 1.111, 95% CI, 1.010—1.221; p = 0.030) remained an independent in-hospital mortality risk factor.

Table 4 Comparison of baseline characteristic between in-hospital non-survivors and survivors among PPF patients experiencing an AE

Table 5 Risk factors for in-hospital mortality after admission in PPF patients who experience an AE

The median survival of patients with an AE after hospitalization was 10.0 months (95% confidence interval [CI], 3.8—16.2 months). Patients with an idiopathic AE did not exhibit survival differences following hospitalization compared to those with triggered AEs (the median survival: 15.0 months vs. 9.0 months, p = 0.305) (Fig. 3B).

Discussion

To our knowledge, this is the first study investigating the AE incidence and clinical outcomes among real-world patients with PPF. During the follow-up period of approximately 3 years, a third of patients experienced an AE, and the 1- and 5-year cumulative AE incidence rates were 12.5% and 38.0%, respectively. AE risk factors included older age, RA-ILD, fibrotic HP, as well as lower DLco. Furthermore, AEs were independently associated with the overall mortality in patients with PPF, along with older age, lower DLco, and a UIP-like pattern on HRCT. During hospitalization, older age, and a lower PF ratio at admission were associated with mortality in patients with AE.

Several previous studies reported that an AE can occur not only in IPF, but also in non-IPF fibrosing ILD. Our study demonstrated a higher AE incidence rate than previous studies [9, 10, 24, 27, 28]. A study by Suzuki et al., including 557 patients with non-IPF fibrosing ILD, reported that AE occurred in 12.4% of patients with non-IPF fibrosing ILD [10]. Kwon et al. conducted a study involving 310 patients with RA-ILD and reported that AEs occurred in 28% of patients over a period of approximately 4 years. Furthermore, the 1- and 3-year AE incidence rates in this study were 9.2% and 19.8%, respectively [24]. A study conducted by Park et al. included 74 and 93 patients with idiopathic NSIP as well as CTD-ILD and reported a 1-year AE incidence rate of 4.2% and 3.3%, respectively [28]. Kang et al. conducted a study involving 101 patients with fibrotic HP, reporting 1- and 3-year AE incidence rates of 6.0% and 13.6%, respectively [9]. The higher AE incidence rates (1- and 5-year incidence rates, 12.5% and 38.0%, respectively) in our study might be explained by including patients who already experienced disease progression despite two years of management.

In our study, older age, RA-ILD, fibrotic HP, and lower DLco were associated with AE development in patients with PPF. Whether older age is an AE-ILD development risk factor is controversial. However, several retrospective CTD-ILD studies found that older age was associated with AE development [12, 24, 29]. Suda et al. conducted a study involving 83 patients with CTD-ILD and reported that older age was significantly associated with AE development (adjusted HR, 1.221; 95% CI 1.054—1.495; p = 0.0052) [12]. Additionally, in a study by Kwon et al. including patients with RA-ILD (n = 310), older age was an AE risk factor in the unadjusted analysis (HR, 1.029; 95%CI, 1.006—1.051, p = 0.013) [24]. Hozumi et al. conducted a study involving 51 patients with RA-ILD and similarly reported that older age at ILD diagnosis was associated with AE development in univariate analysis (HR, 1.03; 95%CI, 1.02—1.21; p = 0.01) [29]. Our study revealed that lower DLco was another AE risk factor consistent with previous studies [9, 10]. Kang et al. conducted a study involving 101 patients with HP and reported that DLco was a significant AE risk factor (adjusted HR, 0.960; 95% CI, 0.939—0.985, p = 0.002) [9]. Similarly, Suzuki et al. reported that lower DLco was significantly associated with AE in a study of 557 patients with non-IPF fibrotic ILD (adjusted HR, 0.876; 95% CI, 0.771—0.995, p = 0.041) [10]. These findings are consistent with previous studies demonstrating that lower lung function is an AE-ILD risk factor, suggesting that patients with more advanced ILD have a higher AE risk than those less advanced [30,31,32].

Although limited data exists on the association between the fibrotic ILD type and the AE risk in non-IPF ILD cases, RA-ILD cases are more likely to develop an AE, as found in several retrospective studies [12, 28]. These findings are similar to our study. In a study by Suda et al. involving 83 patients with CTD-ILD (RA = 25, SJS = 17, SSc = 13, Others = 28), RA-ILD diagnoses were significantly associated with AE development (HR, 2.536; 95% CI, 1.006—11.16, p = 0.0484) [12]. Similarly, Park et al. conducted a study including 74 and 93 patients with idiopathic NSIP and CTD-ILD. They reported that four patients with CTD-ILD experienced an AE, and 75% were RA-ILD cases [28]. Furthermore, fibrotic HP was another independent risk factor for AE development in fibrotic ILD patients in the present study. There is a still lack of data regarding the impact of fibrotic HP on the AE risk of fibrotic ILD. In the aforementioned study by Kang et al., the 1-, 3-, and 5-year cumulative incidence rates of AE in HP were 6.0, 13.6, and 22.8% respectively, suggesting that AE was not uncommon in fibrotic HP patients [9]. Taken together with our findings, fibrotic HP appears to be one of the factors contributing to the development of AEs in patients with fibrotic ILD.

Our study revealed that AEs had a crucial impact on overall survival in patients with PPF. In previous studies, AEs significantly affected the survival of patients with IPF [4, 33,34,35] and those with fibrosing ILDs other than IPF [9, 10, 24, 30]. Alhamad et al. conducted a study including 463 and 204 patients with non-IPF ILD and IPF, respectively. They found that IPF patients and non-IPF ILD patients in the AE groups demonstrated poorer overall survival than those in the non-AE groups (5 years survival rates; 28% [IPF-AE], 50% [non-IPF ILD-AE], 70% [ILD without AE], log-rank test, p < 0.0001) [30]. Suzuki et al. studied patients with non-IPF fibrosing ILD (n = 557). They reported that patients with AE exhibited poorer outcomes than those without (median survival time; 4.3 years vs. not reached, p < 0.001) [10]. In a study by Kang et al. involving 101 patients with fibrotic HP, patients with an AE demonstrated shorter median survival from diagnosis than those without (median survival time; 26 months vs. not reached, p < 0.001) [9]. Considering these findings, further caution and close monitoring are warranted for patients with PPF experiencing AEs, as they may have a poor prognosis.

In our study, older age and lower PF ratios at admission for an AE were significant in-hospital mortality risk factors, consistent with previous reports [9, 10]. In a study by Suzuki et al., a lower PF ratio during an AE diagnosis was an independent predictive factor of 90-day mortality (HR, 0.998; 95% CI, 0.996–1.000, p = 0.041) [10]. Similarly, Kang et al. revealed that a lower PF ratio measured at hospitalization following an AE was significantly associated with in-hospital mortality (adjusted odds ratio, 0.983; 95% CI, 0.967–0.999; p = 0.040) [9].

In our study, non-survivors were more likely to receive steroid pulse therapy than survivors. This is likely to reflect more severe respiratory impairment and greater lung inflammation in these patients, as steroid pulse therapy was given to those with the extensive ground-glass opacities observed on CT scans and higher oxygen requirements. Although the survival benefit of steroid pulse therapy in acute exacerbations of IPF has not been clearly established, corticosteroid treatment remains a recommended intervention according to recent guidelines [7]. Given the clinical and pathobiological similarities between IPF and PPF, it is plausible to consider that treatment strategies for these conditions may be similar. However, further research is needed to determine the efficacy of steroid pulse therapy specifically for acute exacerbations of PPF.

This study has some limitations. First, this study was a retrospective cohort study conducted at a single center, so there may be limitations such as selection bias, recall bias, or lack of generalisability. However, baseline characteristics of our cohort were comparable to those of other studies [26, 36]. Second, we did not use the PPF definition suggested in the international guidelines [3]. However, previous studies suggested that the criteria used in different studies identified populations whose disease progressed similarly [3, 26, 37,38,39]. Further research is needed to validate these findings in patients with PPF based on the new guidelines. Third, patients with unclassifiable ILD were not included in our study. However, unclassifiable ILD is established when specific ILD subtypes cannot be confidently classified despite coordinated efforts by multidisciplinary expertise with input from pulmonologists, thoracic radiologists, and lung pathologists [40]. Therefore, patients with unclassifiable ILD have a heterogeneous population and clinical trajectory, which may cause inconsistent outcomes.

Conclusion

In conclusion, our results suggest that AE is not uncommon and exerts significant prognostic effects for patients with PPF, similar to IPF. Older age, RA-ILD, fibrotic HP, and lower DLco are associated with a higher AE risk in patients with PPF.