Article Text
PDF
Abstract
Background Mycobacterium avium complex (MAC) causes chronic respiratory infectious diseases with diverse clinical features and prognoses. Pleuroparenchymal fibroelastosis (PPFE) is a rare disease characterised by pleural fibrosis with subjacent intra-alveolar fibrosis and alveolar septal elastosis, with unique chest high-resolution CT (HRCT) features (radiological PPFE). An association between recurrent respiratory infections and PPFE formation has been hypothesised; however, the clinical significance of PPFE in MAC lung disease remains unclear.
Methods This retrospective, multicentre study investigated the prevalence of radiological PPFE in patients with MAC lung disease and its association with clinical features and outcomes. Radiological PPFE was diagnosed on the basis of HRCT findings. Prognostic factors were identified using Cox proportional hazards and Fine-Gray models.
Results Of 850 consecutive patients with definite MAC lung disease, 101 (11.9%) exhibited radiological PPFE. Patients with radiological PPFE had unique characteristics, such as lower body mass index, lower survival rate (5-year cumulative survival rate, 63.1% vs 91.7%; p<0.001) and a higher incidence of respiratory-related death (5-year cumulative incidence, 31.1% vs 3.6%; p<0.001), than those without radiological PPFE. In the multivariable analysis, the presence of radiological PPFE was independently associated with all-cause mortality (adjusted HR, 4.78; 95% CI, 2.87 to 7.95; p<0.001) and respiratory-related death (adjusted HR, 3.88; 95% CI, 2.14 to 7.01; p<0.001).
Interpretation This large-scale study demonstrated that in patients with MAC lung disease, radiological PPFE was common, a phenotype associated with unique clinical features and poor prognosis, particularly respiratory-related death. The specific management of this subgroup should be established.
- Interstitial Fibrosis
- Rare lung diseases
- Respiratory Infection
Data availability statement
Data are available upon reasonable request.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
In patients with Mycobacterium avium complex (MAC) lung disease, the prevalence of radiological pleuroparenchymal fibroelastosis (PPFE), a unique chest high-resolution CT feature seen in a rare disease PPFE, and its association with clinical features and outcomes remain unknown.
WHAT THIS STUDY ADDS
Radiological PPFE was detected in approximately 10%–12% of patients with MAC lung disease.
The presence of radiological PPFE was independently associated with lower body mass index, lower survival rate and a higher incidence of respiratory-related deaths.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
This large-scale study demonstrated that radiological PPFE, a phenotype that is associated with unique clinical features and poor prognosis, particularly in respiratory-related deaths such as chronic respiratory failure and pneumonia, were common in patients with MAC lung disease.
For patients with MAC lung disease who have radiological PPFE, multidisciplinary management, including earlier and more aggressive antimycobacterial therapy; pneumonia prevention, including vaccinations, rehabilitation and nutritional therapy; and establishment of specific therapies should be discussed to reduce mortality, particularly respiratory-related deaths resulting from chronic respiratory failure and pneumonia.
Introduction
Non-tuberculous mycobacteria (NTM) are ubiquitous environmental bacteria, some of which are pathogenic to humans.1 Among NTM, the Mycobacterium avium complex (MAC) is considered the major pathogen of lung diseases.2 MAC lung disease has a natural history characterised by progressive respiratory symptoms and is associated with reduced quality of life and poor prognosis. However, the disease course is highly variable and unpredictable3; some patients remain stable for a long time without treatment, while others experience disease progression that leads to fatal outcomes.4 5 Recently, the number of patients with MAC lung disease has been increasing worldwide, making the disease a serious health problem.6–9 Therefore, identifying patients at high risk of death and establishing effective management are urgently needed.
Pleuroparenchymal fibroelastosis (PPFE) is a rare clinical pathological disease entity characterised by pleural fibrosis with subjacent intra-alveolar fibrosis and alveolar septal elastosis associated with poor prognosis.10 PPFE can present as an idiopathic interstitial pneumonia (IIP), but cases secondary to post-lung/bone marrow transplantation and autoimmune diseases have also been reported.10–12 Patients with PPFE have unique chest high-resolution CT (HRCT) findings characterised by dense subpleural consolidation with traction bronchiectasis and architectural distortion, primarily in the upper lobes of the lungs.13–15 Such HRCT features, which are radiological PPFE, coexisted in patients with an IIP other than idiopathic PPFE and those with various pulmonary diseases, including hypersensitivity pneumonitis.10 16–18 An association between PPFE and recurrent respiratory infections has also been suggested; however, evidence regarding the association between PPPE and MAC lung disease is insufficient.16 19 We conducted this large-scale multicentre retrospective study to investigate the prevalence and associated mortality of radiological PPFE in patients with MAC lung disease.
Methods
Patients and diagnostic criteria
We retrospectively included 1299 consecutive patients with MAC lung disease who were diagnosed between 1 January 2006 and 31 August 2020 at Hamamatsu University Hospital, National Hospital Organization Tenryu Hospital, Seirei Mikatahara General Hospital, Seirei Hamamatsu General Hospital or Iwata city Hospital, and had available chest HRCT results (figure 1). For this study, the diagnosis of MAC lung disease was reassessed on the basis of the official American Thoracic Society/European Respiratory Society/European Society of Clinical Microbiology and Infectious Diseases/Infectious Diseases Society of America clinical practice guideline.20 Patients with inadequate available clinical/radiological data (eg, time interval of >6 months between the dates of NTM diagnosis and HRCT, or the apex of the lung not shown on the HRCT images) or those who did not meet the diagnostic criteria were excluded. We identified 850 patients with definite MAC lung disease to estimate the prevalence of radiological PPFE among these patients and determine factors associated with the presence of radiological PPFE. Subsequently, patients with advanced malignancy at MAC diagnosis were excluded. As a result, 818 patients were enrolled in the study cohort to estimate the prognostic significance of radiological PPFE. Previous studies reported that approximately 20%–80% of patients with idiopathic PPFE also had findings of interstitial lung disease (ILD), such as reticular opacities in the lobes other than the upper lobes.14 19 21–25 To minimise the impact of coincidental coexistence of idiopathic PPFE and/or ILD in patients with MAC lung disease and evaluate the prognostic significance of radiological PPFE, we differentiated patients with ILD in lobes other than the upper lobes (ie, patients already diagnosed with ILD before the diagnosis of MAC lung disease or patients with ILD findings confirmed by chest HRCT) from patients without ILD, and we focused on patients without ILD. Moreover, patients with ILD were subanalysed.
Study flow diagram. HRCT, high-resolution CT; ILD, interstitial lung disease; MAC, Mycobacterium avium complex; PPFE, pleuroparenchymal fibroelastosis.
Data collection
Clinical data at the time of diagnosis, including age, sex, body mass index (BMI), smoking status, bacterial species and serological laboratory findings; chest HRCT findings at the time of diagnosis; treatment for MAC lung disease and outcomes, were collected from medical records. Sputum smear scores were assessed semiquantitatively by five scales of –, ±, +, ++ and +++, corresponding to Gaffky scores of 0, 1, 2, 3–6 and 7–10, respectively. Antimicrobial treatment for MAC lung disease was based on the guidelines at that time (eg, clarithromycin/azithromycin, rifampin/rifabutin and ethambutol with/without amikacin or streptomycin).26 The observation period was calculated from the date of diagnosis of MAC lung disease until the last visit (date of censoring or death). Patients who remained alive until 31 August 2020 were censored.
Assessment of chest HRCT images
HRCT findings were determined by a consensus between a pulmonologist and a thoracic radiologist who were blinded to the clinical information. Any disagreement in radiological interpretation was resolved through consensus following discussion. Pulmonary apical cap (PAC) was defined as a homogeneous soft tissue attenuation capping the extreme lung apex (unilaterally or bilaterally), with a sharp or irregular lower border, localised within 1 cm from the apex of the lung on HRCT (figure 2A).27 Radiological PPFE was defined as bilateral, upper lobe and subpleural dense consolidations with traction bronchiectasis below the area within 1 cm from the apex of the lung on HRCT, with a minimum width of 1 cm in contact with the pleura, and distinguished from PAC (figure 2B).12 18 21 28 The HRCT pattern of MAC lung disease was classified as nodular bronchiectatic, fibrocavitary or indeterminate.2 26 ILD findings were defined as bilateral reticular opacities with/without honeycombing on chest HRCT.
Representative images. (A) A representative high-resolution CT (HRCT) image of pulmonary apical cap (arrows). A homogeneous soft tissue attenuation is localised within 1 cm from the apex of the lung. (B) A representative HRCT image of pleuroparenchymal fibroelastosis-like findings (arrowheads). Subpleural dense consolidations with traction bronchiectasis can be observed beyond the area within 1 cm of the apex of the lung.
Statistical analysis
Continuous and categorical variables were expressed as mean±SD or median (IQR) and as number (percentage), respectively. The Welch unequal variances t-test and Fisher exact test or χ2 test were used for between-group comparisons. The cumulative survival rate and incidence of respiratory-related death were evaluated using the Kaplan-Meier method. Between-group differences were assessed using the log-rank test for cumulative survival and Gray’s test for the cumulative incidence of respiratory-related death (treating death from causes other than respiratory-related death as a competing event). Modified Poisson regression analysis was performed to identify factors associated with radiological PPFE; thereafter, risk ratio, 95%CI and p values were calculated. All the variables were included in the multivariable model. Cox proportional hazards regression analysis was used to identify prognostic factors for all-cause mortality, and subdistribution hazard analysis was performed using Fine and Gray’s method to identify the risk factors of respiratory-related death, with death from causes other than respiratory-related death as a competing event. Thereafter, HR, 95% CI and p values were calculated. To assess the independent association of radiological PPFE with outcomes, clinically important variables (eg, already-reported prognostic factors) and treatment exposure (a time-dependent covariate) were included in the univariable and multivariable models.29 30 If any variables showed a correlation coefficient ≥|0.7| between multivariables, the multivariable models were determined by considering the clinical significance and the number of missing data to avoid multicollinearity. Propensity score matching was performed to compare age-matched and sex-matched patients with and those without radiological PPFE (online supplemental method). All statistical analyses were performed using the EZR V.1.36 (Saitama Medical Center, Jichi Medical University, Saitama, Japan) and JMP V.13.2.1 software (SAS Institute, North Carolina, USA). In all analyses, p values of <0.05 were considered statistically significant. The Holm method was used to adjust p values in multiple comparisons.
Supplemental material
Results
Definite MAC lung disease cohort
The characteristics and survival curve of the definite MAC lung disease cohort are shown in online supplemental table S1 and figure S1, respectively. Of the 850 patients, 101 (11.9%) had radiological PPFE. The 5-year cumulative survival rate was 82.8% (95% CI, 79.5% to 85.6%). No patient had macrolide-resistant MAC. Before the diagnosis of MAC lung disease, one patient underwent bone marrow transplantation but did not have radiological PPFE. No patient had a history of lung transplantation. Among the patients with advanced malignancy at MAC diagnosis, 16 underwent prior chemotherapy or chest radiation therapy; of these, 5 had radiological PPFE (online supplemental table 2). Multivariable analysis showed that older age and lower BMI were independently associated with the presence of radiological PPFE (online supplemental table S3).
Study cohort
The characteristics of 818 patients with and without radiological PPFE in the study cohort are presented in table 1. The cumulative survival rate in patients with radiological PPFE was significantly lower than that in patients without radiological PPFE (p<0.001; figure 3A). We observed that the proportion of respiratory-related deaths was higher in patients with radiological PPFE than in those without radiological PPFE (44.4% vs 5.6%). Moreover, the cumulative incidence rate of respiratory-related deaths was significantly higher in patients with radiological PPFE (p<0.001; figure 3B).
Baseline characteristics, treatment and outcomes of patients with radiological PPFE and those without radiological PPFE in the study cohort
Kaplan-Meier survival curves. (A) The 5-year cumulative survival rates of patients with and without radiological PPFE were 56.1% (95% CI, 44.3% to 66.3%) and 89.4% (95% CI, 86.2% to 91.8%), respectively (p<0.001 using the log-rank test). (B) The 5-year cumulative incidence rates of respiratory-related death in patients with and without radiological PPFE were 37.2% (95% CI, 26.4% to 48.0%) and 5.9% (95% CI, 4.0% to 8.3%), respectively (p<0.001 using the Gray’s test). (C) The 5-year cumulative survival rates of PPFE alone and control subjects were 63.1% (95% CI, 49.8% to 73.8%) and 91.7% (95% CI, 88.7% to 94.0%), respectively (p<0.001 using the log-rank test). (D) The 5-year cumulative incidence rates of respiratory-related deaths in PPFE alone and control subjects were 31.1% (95% CI, 19.9% to 43.0%) and 3.6% (95% CI, 2.1% to 5.7%), respectively (p<0.001 using the Gray’s test). PPFE, pleuroparenchymal fibroelastosis;.
The characteristics of patients with ILD and those without ILD in the study cohort are presented in online supplemental table S4. Comorbid ILDs in patients with MAC lung disease included various ILDs, such as IIPs and connective tissue disease (CTD)-associated ILDs. Patients with ILD differed substantially from those without ILD in characteristics and antimicrobial treatment. The prevalence of radiological PPFE was higher in patients with ILD than in those without ILD (24.7% vs 10.2%, respectively; p<0.001). The cumulative survival rate was significantly lower in patients with ILD (p<0.001; online supplemental figure S2).
Patients with MAC lung disease without ILD
The characteristics of patients without ILD in the study cohort are summarised in table 2. Of the 725 patients, 74 (10.2%) had radiological PPFE. The patients without ILD but with radiological PPFE (PPFE alone) were older, had a higher proportion of men, had lower BMIs, lower serum albumin levels and higher C reactive protein levels than those without ILD or radiological PPFE (control subjects). As for the causative bacterial species, the prevalence of M. intracellulare was significantly higher in PPFE alone than in control subjects. The cumulative survival rate was significantly lower in PPFE alone than in control subjects (p<0.001; figure 3C). The cumulative incidence rate of respiratory-related deaths was significantly higher in PPFE alone (p<0.001; figure 3D). No significant between-group difference in the cumulative incidence of pneumothorax requiring hospitalisation was found (p=0.282 by the Gray’s test). For a sensitivity analysis, we compared age-matched and sex-matched propensity score-matched PPFE alone and control subjects (online supplemental table S5 and figure S3). The matched PPFE alone had significantly lower BMIs and serum albumin levels than the matched control subjects, but the between-group differences in the other characteristics were not significant. Even after propensity score matching, PPFE alone had a lower cumulative survival rate (p<0.001; online supplemental figure S4a) and higher incidence rate of respiratory-related death than control subjects (p<0.001; online supplemental figure S4b).
Baseline characteristics, treatment and outcomes of patients without interstitial lung disease
Control subjects were classified into those with and without PAC, and then compared (online supplemental table S6). The patients with PAC were older and had lower BMIs than those without PAC, but the between-group differences in the other characteristics were not significant. The between-group difference in cumulative survival rate was also not significant (online supplemental table S5).
Patients with MAC lung disease with ILD
The characteristics of patients with ILD in the study cohort are summarised in (online supplemental table S7). Patients with ILD who had radiological PPFE (PPFE-and-ILD) had a significantly lower BMI than those without radiological PPFE (ILD alone). The cumulative survival rate was significantly lower in PPFE-and-ILD than in ILD alone (p<0.001; online supplemental figure S6a). The cumulative incidence rate of respiratory-related deaths was also significantly higher in PPFE-and-ILD (p=0.012; online supplemental figure S6b).
Comparison of PPFE alone and PPFE-and-ILD
The characteristics of PPFE alone and PPFE-and-ILD are demonstrated in online supplemental table S8. The cumulative survival rate of PPFE-and-ILD was significantly lower than that of PPFE alone (p=0.0057; online supplemental figure S7a). Moreover, the cumulative incidence rate of respiratory-related deaths was significantly higher in PPFE-and-ILD than in PPFE alone (p=0.043; online supplemental figure S7b).
Prognostic significance of radiological PPFE
The results of the Cox proportional hazard analysis of all-cause mortality in the study cohort are presented in table 3. The multivariable analysis revealed that the presence of radiological PPFE was an independent prognostic factor for all-cause mortality (adjusted HR, 4.78; 95% CI, 2.87 to 7.95; p<0.001).
Cox proportional hazards regression analysis of all-cause mortality in the study cohort
The results of the subdistribution hazards analysis of respiratory-related death in the study cohort are presented in table 4. The multivariable analysis revealed that ILD comorbidity (adjusted HR, 2.50; 95% CI, 1.16 to 5.39; p=0.020) and the presence of radiological PPFE (adjusted HR, 3.88; 95% CI, 2.14 to 7.01; p<0.001) were independent risk factors of respiratory-related death.
Fine-Gray subdistribution hazards analysis for respiratory-related death
Discussion
In this multicentre retrospective study, the prevalence of radiological PPFE in the definite MAC lung disease cohort was 11.9%. In the study cohort excluding patients with advanced malignancy, the presence of radiological PPFE was an independent prognostic factor for all-cause mortality. We also found that the presence of radiological PPFE was independently associated with an increased risk of respiratory-related death. However, the presence of PAC was not associated with prognosis in the patients with MAC lung disease. To the best of our knowledge, this is the largest study to investigate the prevalence and prognostic significance of radiological PPFE and the first study to demonstrate their independent associations with respiratory-related death in patients with MAC lung disease.
The prevalence of idiopathic PPFE in the general population is unknown. The reported prevalence of idiopathic pulmonary fibrosis (IPF), a major IIP, ranges from 0.5 to 27.9 per 100 000 people.31 32 However, the prevalence of idiopathic PPFE, a rare IIP, can be estimated to be lower than that of IPF. The prevalence of radiological PPFE in the general population is also unknown. Studies have shown that radiological PPFE was found in 6.3% of patients with IPF,33 18% of patients with scleroderma34 and 40% of patients with hypersensitivity pneumonitis.35 In this study, none of the patients with radiological PPFE had scleroderma or hypersensitivity pneumonitis, and few of them had CTD. For NTM, a single-centre retrospective study reported that radiological PPFE was found in 59 (26.3%) of 224 patients with MAC lung disease.16 Our multicentre study involving 850 patients showed that the prevalence of radiological PPFE was 11.9%, demonstrating that radiological PPFE is a common HRCT feature in MAC lung disease. Given that several reports suggest an association between recurrent infections and PPFE formation and that the prevalence of radiological PPFE in patients with MAC lung disease was relatively high despite the very low estimated prevalence of idiopathic PPFE, we may be able to hypothesise that MAC infection promotes the development of PPFE (ie, PPFE secondary to MAC) or that patients with radiological PPFE are more susceptible to MAC infection.
Older age, male sex, lower BMI, lower albuminemia and a fibrocavitary-type finding on HRCT are factors associated with poor prognosis in NTM disease,29 30 consistent with the results of our study. In this study, the presence of radiological PPFE was an independent prognostic factor in patients with MAC lung disease, even after adjustment for these already-reported prognostic factors, the presence of ILD comorbidity and treatment exposure. Similar results were reported in a single-centre retrospective study,16 but the study had statistical limitations owing to its small sample size and results based on a logistic analysis of a binary variable of whether the patients died within 5 years, rather than a time-to-event model (ie, Cox proportional hazards model). Additionally, the study did not report data on cumulative survival or cause of death. By contrast, our large-scale multicentre study showed that the 5-year cumulative survival rate was significantly lower in the patients with than in those without radiological PPFE. Similar results were replicated in the patients who were propensity score matched for age and sex. In addition, our Cox proportional hazard analysis results demonstrated that the presence of radiological PPFE was an independent prognostic factor. In the patient group with radiological PPFE, the incidence rate of respiratory-related death was significantly higher than that in the group without radiological PPFE, and Fine-Gray analysis results demonstrated that radiological PPFE was independently associated with respiratory-related deaths. Collectively, the patients with MAC lung disease with radiological PPFE represented a poor prognostic phenotype associated with respiratory-related death. Specific treatment strategies for such patients need to be established.
In this study, respiratory-related deaths caused by chronic respiratory failure and pneumonia were common in patients with radiological PPFE. Therefore, to improve mortality in these patients, treatment strategies to prevent chronic respiratory failure and pneumonia are needed. One of the causes of chronic respiratory failure was the progression of MAC lung disease. In addition, recurrent/chronic respiratory infections due to MAC could have induced the development/progression of PPFE, as suggested in previous studies.16 19 Thus, earlier and more aggressive antimycobacterial therapy, including those using aminoglycosides, should be considered for these patients.2 Further, the hypothesis that radiological PPFE is similar in pathogenesis to idiopathic/secondary PPFE can be considered. However, there is still insufficient evidence on pharmacological treatment for idiopathic/secondary PPFE, including antifibrotic agents such as nintedanib and pirfenidone as well as corticosteroids.36 37 To reduce pneumonia-related deaths, pneumonia prevention strategies,38 39 including pneumococcal and influenza vaccinations, should be considered. Additionally, although there is no specific treatment to prevent each cause of death, rehabilitation and nutritional therapy have been reported to improve the quality of life and prognosis of patients with chronic respiratory diseases, such as bronchiectasis, chronic obstructive pulmonary disease and IPF.40–42 Therefore, these multidisciplinary treatments should be discussed in patients with MAC lung disease who have radiological PPFE.
An association between idiopathic PPPE and low BMI has been reported. In IPF and collagen diseases such as scleroderma, patients with radiological PPFE had lower BMIs than those without radiological PPFE.12 33 34 43 Similar findings were observed in patients with MAC lung disease, as shown in this study. In addition, our multivariable analysis revealed an independent association between radiological PPFE and low BMI in patients with MAC lung disease. Whether the progression of radiological PPFE promotes lower BMI, whether low BMI induces radiological PPFE in the lung, or whether patients with low BMIs have a predisposition to developing radiological PPFE is unclear. Some common mechanisms might exist across diseases for the association between radiological PPFE and low BMI.
Radiological PPFE has also been reported to be associated with poor prognosis in IPF and collagen diseases, such as scleroderma.12 18 33 As mentioned earlier, the presence of radiological PPFE was associated with low BMI. Reportedly, low BMI was associated with poor nutrition and sarcopenia, and a prognostic factor in some respiratory diseases, including IPF.30 44 Therefore, the prognostic relevance of radiological PPFE may be due to their confounding effect on low BMI. However, in this study, as radiological PPFE and low BMI were independently associated with respiratory-related death in the patients with MAC lung disease, other reasons might need to be considered. A study showed that patients with IPF who had radiological PPFE had greater deterioration of pulmonary function over time than those without radiological PPFE.33 We could hypothesise that in patients with MAC lung disease, pulmonary function may deteriorate more rapidly in those with than in those without radiological PPFE, which may be associated with increased respiratory-related death. Prospective studies to evaluate pulmonary function over time in MAC lung disease are needed to prove this hypothesis.
The differential diagnosis for radiological PPFE is PAC.27 28 45–47 PAC is generally confined to the pulmonary apex and is thought to be rarely progressive.48 49 In this study, radiological PPFE and PAC were distinguished by simple and clear HRCT criteria. The characteristics and cumulative survival rates of patients with PAC and those without PAC were not significantly different, but rather appeared to be comparable. This may support the possibility that radiological PPFE may be a different condition from PAC. However, because we did not evaluate changes in chest HRCT findings over time in this study, we cannot rule out the possibility that cases of early PPFE were included in the PAC group. Prospective studies with regular follow-up using HRCT are needed to address this issue.
ILDs comprise various diseases, each with a different prognosis, but their presence/comorbidity is often linked to death.50 51 Therefore, in this study, the fact that patients with ILD had a poorer prognosis than those without ILD is probably reasonable. We also found that even among patients with ILD, patients with radiological PPFE (PPFE-and-ILD) had a lower survival rate and higher respiratory-related mortality than those without radiological PPFE (ILD alone), demonstrating a strong influence of radiological PPFE on prognosis. Interestingly, compared with PPFE alone, PPFE-and-ILD had slightly different characteristics and a poorer prognosis. PPFE-and-ILD might have included patients with incidental coexistence of MAC lung disease with idiopathic PPFE, patients with incidental coexistence of PPFE secondary to MAC lung disease with an ILD and/or patients with radiological PPFE that have progressed to the lower lobes as ILD, but further research that will focus on this subgroup is required.
Our study has several limitations. First, this was a retrospective study; therefore, potential bias might have affected our results. Second, many patients had no pulmonary function test results. Prospective study of the association of radiological PPFE with lung function and prognosis may yield interesting results. Third, we did not evaluate the histopathology of radiological PPFE. Surgical lung biopsy is often avoided in PPFE because of the recognised risk of iatrogenic refractory complications such pneumothorax, pneumomediastinum or bronchopleural fistula, which could cause a protracted air leak.10 Therefore, several diagnostic criteria based on clinical and radiological findings without pathological evidence, including ours,12 52 have been proposed; moreover, the number of PPFE publications based on these criteria is increasing.34 35 53 However, it may be necessary to evaluate the histopathology to further identify PPFE in MAC lung disease in future studies. Fourth, not all patients regularly underwent repeated HRCT after the diagnosis of MAC lung disease. Therefore, we cannot exclude the probability that subsequent ILD development might have been missed in some patients. Finally, the treatment for MAC lung disease differed among the patients. However, the Cox proportional hazards analysis with time-dependent covariates was used in this study to minimise the influence of differences in treatment initiation timings.
Interpretation
This large-scale multicentre study showed that in patients with MAC lung disease, radiological PPFE was common. The patients had unique clinical features, such as low BMI and poor prognosis, mainly associated with respiratory-related death. A specific management for patients with MAC lung disease presenting with such a phenotype should be established. We believe that the insights from this study will be a keystone for establishing this concept of PPFE in MAC lung disease.
Data availability statement
Data are available upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This multicentre study was conducted in accordance with the Declaration of Helsinki, and the study protocol was approved by the ethics committees of Hamamatsu University School of Medicine (approval number: 20-265) and the other hospitals (approval numbers: 2020-12, 20-61, 3524 and 2020-036). Patient approval or the informed consent requirement was waived because of the retrospective nature of the study.
Acknowledgments
We thank Enago for the English language editing.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors YA and HH: conception and design; data collection, analysis and interpretation; manuscript writing and final approval of the manuscript. MKono, DA, YI, KM, MKarayama, YS and KF: conception and design as well as data collection and analysis. DH, HN, KY, SI, MS, NE, TF, YN and NI: data collection and analysis, and supervision. TS: conception and design, manuscript writing and administrative support. All authors reviewed and confirmed the final manuscript. HH is the guarantor of the article.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.