Article Text
PDF
Abstract
Background Interstitial lung abnormalities (ILAs) are associated with the risk of lung cancer and its mortality. However, the impact of ILA on treatment-related complications and survival in patients who underwent curative surgery is still unknown.
Research question This study aimed to evaluate the significance of the presence of computed tomography-diagnosed ILA and histopathologically matched interstitial abnormalities on postoperative pulmonary complications (PPCs) and the long-term survival of patients who underwent surgical treatment for lung cancer.
Study design and methods A matched case–control study was designed to compare PPCs and mortality among 50 patients with ILA, 50 patients with idiopathic pulmonary fibrosis (IPF) and 200 controls. Cases and controls were matched by sex, age, smoking history, tumour location, the extent of surgery, tumour histology and pathological TNM stage.
Results Compared with the control group, the OR of the prevalence of PPCs increased to 9.56 (95% CI 2.85 to 32.1, p<0.001) in the ILA group and 56.50 (95% CI 17.92 to 178.1, p<0.001) in the IPF group. The 5-year overall survival (OS) rates of the control, ILA and IPF groups were 76% (95% CI 71% to 83%), 52% (95% CI 37% to 74%) and 32% (95% CI 19% to 53%), respectively (log-rank p<0.001). Patients with ILA had better 5-year OS than those with IPF (log-rank p=0.046) but had worse 5-year OS than those in the control group (log-rank p=0.002).
Conclusions The presence of radiological and pathological features of ILA in patients with lung cancer undergoing curative surgery was associated with frequent complications and decreased survival.
- interstitial fibrosis
- lung cancer
- thoracic surgery
Data availability statement
Data are available on reasonable request.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Recent cohort studies have demonstrated that patients with interstitial lung abnormalities (ILAs) are at an increased risk of developing lung cancer and its associated mortality. However, the impact of ILAs with histopathological evidence of interstitial abnormalities in surgical specimens on treatment-related complications and survival in patients with lung cancer who underwent curative surgery is still unknown.
WHAT THIS STUDY ADDS
Patients with ILA are at an increased risk of treatment-related complications and decreased survival compared with otherwise healthy controls, but lower risk of complications and better survival compared with idiopathic pulmonary fibrosis.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE AND/OR POLICY
The adverse impact of ILAs on outcomes of surgical treatment for lung cancer warrants further studies to establish optimal surveillance and treatment strategies for patients with lung cancer with ILAs.
Introduction
Interstitial lung abnormalities (ILAs) refer to incidental chest CT findings of ground glass or reticular abnormalities, traction bronchiectasis and/or bronchiolectasis, honeycombing and/or non-emphysematous cysts involving at least 5% of a lung zone in individuals without a known diagnosis or clinical suspicion of interstitial lung disease (ILD).1 ILAs are increasingly recognised on chest CTs performed for various purposes, such as lung cancer screening.2 3 The reported prevalence of ILA in population-based cohorts and lung cancer screening cohorts ranges from 3% to 17%, being slightly higher in smokers than in non-smokers.4–7 Although the clinical significance of ILA remains to be elucidated, increased all-cause mortality as well as reduced pulmonary function and progression of ILA, possibly into ILD have been demonstrated.6 8 9
The frequent development of lung cancer in idiopathic pulmonary fibrosis (IPF), one of the most common and fatal ILD, is well known. Previous studies on IPF demonstrated a 10-year cumulative incidence of lung cancer in patients with IPF as 31.1%–54.7%.10–12 In addition, patients with lung cancer with concomitant IPF suffer from high rates of treatment-related complications, leading to poor prognosis.11 13–15 In this regard, the significance of ILA in the development and prognosis of lung cancer has been of particular interest.1 Recent studies on lung cancer screening cohorts and population-based cohorts have shown an increased prevalence and incidence of lung cancer in subjects with ILA.7 16 17 Furthermore, the presence of ILA is an independent risk factor for lung cancer mortality.17 However, the precise impact of ILA on treatment-related complications and survival of patients with specific treatment modalities is still unknown. Moreover, as the diagnosis of ILA is solely based on radiologic findings, the significance of the histological features of interstitial abnormalities in lung cancer has not been fully assessed. Notably, previous studies have identified that histological features of ILD, such as interstitial fibrosis or usual interstitial pneumonia from resected lung specimens, are common in patients without clinical or subclinical ILD.18–20
This study aimed to evaluate the impact of CT-diagnosed ILAs being matched with surgical-specimen-based histopathological evidence of interstitial abnormalities on postoperative pulmonary complications (PPCs) and the long-term survival of patients who underwent surgical treatment for lung cancer. A case–control study was conducted to assess whether the PPCs and mortality were significantly different among the control, ILA and IPF groups.
Methods
Study design and recruitment
This was a matched case–control study conducted by the Samsung Medical Center, a tertiary referral centre in Seoul, South Korea (1989 beds, university affiliated) designed to compare PPCs and mortality among patients with ILA, IPF and controls. The routine preoperative evaluation of patients with pulmonary malignancy included pulmonary function tests, chest CT scans and flexible bronchoscopy. Positron emission tomography/CT and brain MRI were also performed to assess distant metastasis of cancer. For case selection, all patients who underwent curative resection for lung cancer between January 2010 and April 2020 and had a chest CT scan within 3 months before surgery were eligible. Patients who received neoadjuvant chemotherapy and/or radiotherapy were excluded because of their potential to elicit ILA or ILD.
Radiologic diagnosis of ILA
As the definition of ILAs is purely radiological and is based on the incidental identification of CT abnormalities, preoperative chest CT studies were retrospectively analysed by two pulmonologists (HY and YI) who reached a consensus on the presence of ILA.1 For cases where the diagnosis was in disagreement, the cases were discussed until consensus was reached. The records of these patients were reviewed for inclusion in the study. Patients were excluded if they had any prediagnosed ILD. Finally, 50 cases of ILA combined with lung cancer were included and the CT features were matched with histopathological specimens obtained after lung cancer surgery.
Histopathological review of specimen
The histopathological diagnosis of ILA was based on the identification of histological features of interstitial abnormalities such as interstitial fibrosis and/or inflammation from pulmonary resection specimens for lung cancer. Pathology slides of ILA cases were reviewed by two board-certified pathologists (JH and SP) who were completely blinded to the chest CT findings and outcomes of the study. Disagreement between the pathologists was resolved by discussion until consensus was reached. Histopathological review of the controls was also performed for the absence of evidence for interstitial abnormalities.
Selection of controls
We identified 200 patients with lung cancer in the control group without any pulmonary disease including chronic obstructive pulmonary disease (COPD) except for lung cancer and 50 patients with IPF and lung cancer who underwent curative resection for lung cancer at the same institution between January 2012 and December 2016, from our prospective registry of IPF10 and a separate prospective registry of patients with lung cancer.21 COPD was defined as the presence of airflow limitation established by post-bronchodilator spirometry in proper clinical context according to the Global Initiative for Chronic Obstructive Lung Disease guideline.22 The diagnosis of IPF was based on the diagnostic criteria of the American Thoracic Society and European Respiratory Society.23 Six variables (sex, age, tumour location, the extent of surgery, tumour histology and pathological TNM stage) were considered as matching variables based on previous reports.24–27 Cases and controls with the same values of matching variables were directly extracted from the database and individual exact match were performed; 1:1:4 of total 50 pairs were matched.
Data collection
For cases and controls, detailed data on clinical characteristics, including age at the time of surgery, sex, smoking history, pulmonary function test, tumour location and histology, pathological stage (staged or restaged according to the eighth edition of the TNM staging system), and the extent of surgery were obtained from electronic medical records. All dates of patient deaths were extracted from the electronic medical records database, ascertained from in-hospital medical records or the database of the National Health Insurance Service. The obtainment of the data was approved by the IRB.
Definition of PPC
We considered the following PPCs that occurred during 60 postoperative days: pneumonia, acute lung injury, respiratory failure, significant atelectasis requiring bronchoscopy or reintubation, bronchopleural fistula/empyema, prolonged air leakage lasting for more than 5 days, pneumothorax, pleural effusion and pulmonary thromboembolism.28 29 PPCs with grade 2 or higher based on the Clavien‐Dindo classification of surgical complications were considered significant.
Statistical analysis
Data are presented as median (IQR) or number (n) (%). Categorical variables were compared using the Pearson’s χ2 test, and continuous variables were compared using the Kruskal-Wallis test. Overall survival (OS) rate was estimated using the Kaplan-Meier method and adjusted using the Benjamini-Hochberg method, and survival distributions between cases and controls were compared using the log-rank test.
Univariable and multivariable logistic regression and Cox proportional hazards regression analyses were performed to identify factors associated with PPCs or death. Variables with known clinical importance were included in the multivariable analyses. The results were reported as ORs or HRs with 95% CIs. Schoenfeld’s global test was used to test the proportional hazards assumption in the Cox proportional hazards model, focusing on whether the effects of covariates on risk remain constant over time. All tests were two sided, and a p<0.05 was considered statistically significant. All statistical analyses were performed using R (V.3.5.3; R Foundation for Statistical Computing, http://www.r-project.org). ‘Survival’ package in R software was used to perform the analysis.
Results
Baseline characteristics
The median (IQR) age of the patients was 68 (62–72) years, 89% were males, and 84% were former or current smokers (table 1). Regarding tumour histology, adenocarcinoma was the most common (n=177, 59%), followed by squamous cell carcinoma (n=117, 39%). Most patients underwent lobectomy (n=254, 85%). Chest CT scan and matched pathology are shown in online supplemental figure 1s.
Supplemental material
Baseline characteristics of study subjects according to three groups: control, patients with ILA and patients with IPF
Baseline characteristics of the control, ILA and IPF groups
A total of 200 otherwise healthy patients without any pulmonary disease except for lung cancer and 50 patients with IPF were matched with 50 patients with ILA (table 1, online supplemental figure s2). Cases and controls were not statistically significantly different in terms of age, sex, smoking history, smoking intensity and pulmonary function, including forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC) and FEV1/FVC. However, a trend of decreasing diffusing capacity for carbon monoxide was observed in the control, ILA and IPF groups (17.9 vs 14.7 vs 14.0, p<0.001). There were no statistically significant differences in tumour location and histology, the extent of surgery and pathological TNM stage among the three groups.
PPCs and survival of the control, ILA and IPF groups
During the study period, 39 (13%) patients developed PPCs following pulmonary resection (table 1). Among these patients, the most common type of PPC was prolonged air leakage (n=19, 49%), followed by acute lung injury (n=12, 31%), pneumonia (n=6, 16%) and pneumothorax (n=2, 0.5%). A significant increase in the prevalence of PPCs was observed in the control, ILA and IPF groups (3% vs 18% vs 50%, p<0.001) (table 1).
A total of 88 (29%) patients died during the median follow-up of 52.9 (28.3–60.0) months. Of the 50 patients with ILA, 16 (32%) patients died during the median follow-up of 28.4 months (IQR, 17.9–44.9) months. Of the 50 patients with IPF, 26 (52%) patients died during the median follow-up of 20.4 (IQR 12.7–43.3) months and 46 (23%) out of the 200 controls died during the median follow-up 56.4 (IQR 51.0–62.6) months (table 1). The 5-year OS rates of the control, ILA and IPF groups were 76% (95% CI 71% to 83%), 52% (95% CI 37% to 74%) and 32% (95% CI 19% to 53%), respectively (log-rank p<0.001). Patients with ILA had better 5-year OS rates than those with IPF (log-rank p=0.046) but had worse 5-year OS rates than the controls (log-rank p=0.002) (figure 1). Based on the Schoenfeld residual test, patients in the control, ILA and IPF groups met the proportional hazard assumption with p=0.716.
Kaplan-Meier survival plots of the study subjects according to the three groups: controls, patients with interstitial lung abnormalities (ILA) and patients with idiopathic pulmonary fibrosis (IPF).
Risk factors of PPCs and mortality
A multivariable logistic regression model for evaluating risk factors for the development of PPCs is shown in table 2. Among the potential risk factors, age (adjusted OR 1.12; 95% CI 1.04 to 1.20; p=0.004), the presence of ILA (adjusted OR 9.56; 95% CI 2.85 to 32.1; p<0.001) or IPF (adjusted OR 56.50; 95% CI 17.92 to 178.1; p<0.001) and lobectomy (adjusted OR 4.46; 95% CI 1.11 to 17.9; p=0.035) were independently associated with the occurrence of PPCs. A multivariable logistic regression using clustering within matched pairs showed similar results of age and presence of ILA and IPF being associated with the occurrence of PPCs (online supplemental table s1).
Univariable and multivariable logistic regression analysis for factors related to PPCs in study patients (n=300)
Univariable and multivariable Cox analyses for factors related to 5-year mortality in the study patients were conducted (table 3). The presence of ILA (adjusted HR 2.64; 95% CI 1.47 to 4.74; p=0.001) or IPF (adjusted HR 4.68; 95% CI 2.84 to 7.71; p<0.001) were significantly associated with increased mortality. In multivariable Cox analysis using clustering within matched pairs, the presence of ILA was still associated with increased mortality (online supplemental table s2).
Univariable and multivariable Cox analysis for factors related to 5-year mortality in study patients (n=300)
Histopathological review of specimen
Histopathological features and patterns of patients with ILA are described in online supplemental table s3. Interstitial fibrosis, fibroblastic foci and emphysema were observed in 21 (42%), 12 (24%) and 19 (38%) patients, respectively. Thirteen (26%) patients were classified as usual interstitial pneumonia, while 2 (4%) were diagnosed with smoking-related ILD pattern. Presence of specific features and patterns were not associated with the occurrence of PPCs and mortality (online supplemental table s3, supplemental figures s2 and s3).
Discussion
In this matched case–control study, we assessed the impact of ILA with histopathological evidence of interstitial abnormalities in surgical specimens on PPCs and the long-term survival of patients with lung cancer who underwent curative surgery. Compared with the control group, the OR of the prevalence of PPC increased to 9.56 (95% CI 2.85 to 32.1, p<0.001) in the ILA group and 56.50 (95% CI 17.92 to 178.1, p<0.001) in the IPF group. Similarly, a significant difference in the 5-year OS rates among the three groups was observed, with the 5-year survival rate of ILA being lower than that of the controls but higher than that of IPF (controls vs ILA vs IPF: 76% vs 52% vs 32%). The presence of ILA was an independent risk factor for PPCs and mortality in Cox regression analysis.
To the best of our knowledge, this is the first study to analyse the implications of the presence of pathologically confirmed ILA on survival and treatment-related complications in a specific group of patients with lung cancer who underwent curative surgical treatment. In addition to studies based on population or lung cancer screening cohorts that have demonstrated increased all-cause mortality of patients with lung cancer with ILA, several studies assessed the influence of ILA on treatment-related complications of various treatment modalities. Nakanishi et al reported that pre-existing ILA and ground-glass attenuation were associated with higher rates of immune checkpoint inhibitor-induced pneumonitis after evaluating 83 patients with stage IV non-small cell lung cancer.30 Similarly, patients with lung cancer with ILA are at an increased risk of severe radiation pneumonitis following thoracic radiotherapy.31 32 In a study by Im et al from our institution, ILA was an independent risk factor for PPC in elderly patients (aged ≥70 years) with lung cancer.33 The results of our study indicate that the presence of ILA is associated with a higher risk of PPCs as well as mortality in general patients with lung cancer undergoing surgical treatment. Furthermore, when the rates of PPCs and mortality were compared with those of patients with IPF, the disease well known for its detrimental effect on the outcomes of lung cancer, patients with ILA demonstrated a better prognosis. The adverse impact of ILA on outcomes of surgical treatment for lung cancer warrants further studies to establish optimal surveillance and treatment strategies for patients with lung cancer with ILA.
Among various PPCs, acute lung injury (ALI) was observed in 2 (1%), 4 (8%) and 6 (12%) patients of the control, ILA and IPF groups, respectively. Although the prevalence of ALI following lung resection varies widely, the reported mortality of ALI is uniformly high.34 35 Not only ILD including IPF is a known risk factor of ALI following lung cancer surgery, IPF is well known for its unique disease characteristic of acute exacerbation.36 The prognosis of acute exacerbation is poor, with the in-hospital mortality reaching up to 50%.37 Despite that precise pathogenesis of both postoperative ALI and acute exacerbation of IPF are not elucidated, excessive inflammation induced by certain triggers such as surgery, drugs or aspiration eliciting alveolar epithelial tissue damages are thought to be an important mechanism.36 38 Furthermore, they share similar radiological characteristics of diffuse ground-glass opacities and pathological features of diffuse alveolar damage.39–41 Hence, frequent occurrence of acute exacerbation or ALI following surgery and their detrimental effect on survival in IPF patients with lung cancer poses a great challenge in treatment of these patients. In that matter, the frequency of ALI in patients with ILA has been of interest. Our results indicate that patients with ILA have an intermediate risk of ALI development compared with IPF or otherwise healthy controls. Future research on pathogenesis and optimal treatment for these patients is necessary. Especially, recent studies on IPF and patients with lung cancer suggested that perioperative treatment with antifibrotic agents may be helpful in reducing the development of ALI.42 The role of such agents on patients with ILA and lung cancer requires further investigation.
One of the important aspects of our study that requires attention is that ILA in our study was defined as the presence of both radiological and pathological evidence of interstitial abnormalities. ILA, by definition, is a purely radiological term and is based on the incidental identification of CT abnormalities.2 Accurate identification, characterisation and classification of ILA may vary depending on the CT protocol.43 44 Thus, ILA may encompass heterogeneous conditions, including dependent atelectasis and air trapping, as well as mild forms of pulmonary fibrosis. This heterogeneity of ILA is also portrayed in the diverse prognosis of the disease.3 8 A recent study revealed that certain patterns and distributions may aid in differentiating the subset of patients with risk of progression who will eventually require treatment.8 However, taking the example of ILD, a substantial rate of discordance between radiological and pathological patterns, especially in the non-usual interstitial pneumonia pattern,23 45 and the coexistence of multiple pathological patterns in a single patient or a single biopsy specimen,46 which impose significant impact on treatment and survival,47 have been documented. Thus, the importance of pathological review is advocated in diagnosis and treatment of ILD. Hence, an analysis of pathological patterns and radiological/pathological correlation may provide an insight in understanding the disease of ILA. Experts recommend assessing background lung from cancer resections and documenting histological patterns suspicious for ILD.1 In addition, an incidentally identified histological evidence of interstitial abnormalities is regarded as an area of uncertainty requiring future research. Nonetheless, mild nature of the ILA, spontaneous resolution of disease in some patients, and risk of surgical biopsy in the context of unclear benefit limits the identification of pathological features as well as radiological/pathological correlation. In this study, we demonstrated that a subset of ILA represents histological lesions of usual interstitial pneumonia. Furthermore, we confirmed that ILA accompanied by pathological abnormalities is associated with treatment-related complications and survival in lung cancer. The results of our study suggest the need for additional studies to determine the progression and mortality based on pathological patterns of ILA and the correlation between pathological and radiological features of ILA.
Potential limitations should be acknowledged to fully appreciate the results of our study. First, given the retrospective observational nature, there is always the possibility that selection bias of confounding factors might have influenced our findings. Regarding the possibility that the presence of IPF or ILA may be related to the differences in baseline characteristics or affected the treatment, we tried to minimise this limitation by the case-matching study design. However, this statistical method may not be sufficient to overcome bias. Second, patients with IPF or ILA in our study were ones considered suitable for surgical treatment of lung cancer despite underlying pulmonary disease after careful evaluation. Therefore, the results of our study may not be generalised to other patients with lung cancer with IPF or ILA who are candidates for other treatment modalities. Third, as we defined ILA as the presence of pathological evidence of interstitial abnormalities as well as radiological abnormalities, patients classified as having ILA in our study may have a more severe status of ILA since non-significant patients, such as those with dependent atelectasis were excluded. Thus, it should be noted that the patients with ILA in our study may be a selective homogenous subgroup of ILA which may have affected the clinical outcomes of PPCs or mortality. Furthermore, patients in whom ILA and lung cancer did not coexist in the same lobe, such as cases of ILA in the lower lobe and lung cancer in the upper lobe were also excluded from our study. The significance of ILA in such cases requires further study. Forth, patients with COPD, along with other pulmonary diseases, was excluded in the control group. COPD is a potential risk factor of PPCs following lung cancer surgery, thus the exclusion may have influenced the outcome of our study patients. However, the prevalence of airflow limitation based on prebronchodilator spirometry across the control, ILA and IPF groups did not differ. Furthermore, when airflow limitation was included in the multivariable analysis, it was not associated with PPCs or mortality as well as did not alter the statistical significance of ILA or IPF. Finally, although patients with ILA were at an increased risk of treatment-related complications and decreased survival compared with otherwise healthy controls, direct impact of PPCs on survival could not be determined as cause of death was not available in some patients. As a detrimental influence of treatment-related complications on survival of IPF patients with lung cancer has been documented, further research is necessary to determine the relationship between complications and survival in ILA patients with lung cancer.
Conclusions
The presence of histopathological and radiological features of ILA in patients with lung cancer undergoing curative surgery was associated with frequent complications and decreased survival.
Data availability statement
Data are available on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
The Institutional Review Board of Samsung Medical Center approved the collection, analysis and publication of the data and waived the requirement for informed consent as we only used deidentified data retrieved from electronic medical records (IRB no. 2020-10-070-001).
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
YI and MPC contributed equally.
Contributors Study conception and design: YI and HY. Data acquisition and analysis: YI, MPC, KSL, JH, MJC, HKK, JHC, YSC, SP, HJK, OJK, BP, HY. Data interpretation and manuscript writing: YI and HY. Critical revision and final approval of the manuscript: all authors. HY is responsable for the overall content as the guarantor.
Funding This research was supported by the MSIT (Ministry of Science and ICT), Korea, under the ICT Creative Consilience program (IITP-2021-2020-0-01821) (NTIS 1711126102) supervised by the IITP (Institute for Information and Communications Technology Planning & Evaluation) and SMC-SKKU (Samsung Medical Center-Sungkyunkwan University) grant (SMO1201081).
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.