You are here

  • Home
  • Archive
  • Volume 78, Issue 7
  • Airflow limitation and mortality during cancer screening in the National Lung Screening Trial: why quantifying airflow limitation matters

Article Text

Article menu
Lung cancer
Original research
Airflow limitation and mortality during cancer screening in the National Lung Screening Trial: why quantifying airflow limitation matters
Free
  1. Robert P Young1,
  2. Ralph C Ward2,
  3. Raewyn J Scott1,
  4. Greg D Gamble1,
  5. Gerard Silvestri2
  1. 1 The University of Auckland Faculty of Medical and Health Sciences, Auckland, New Zealand
  2. 2 Medical University of South Carolina, Charleston, South Carolina, USA
  1. Correspondence to Dr Robert P Young, The University of Auckland Faculty of Medical and Health Sciences, Auckland, Auckland, New Zealand; roberty{at}adhb.govt.nz

Abstract

Importance Current eligibility criteria for lung cancer (LC) screening are derived from randomised controlled trials and primarily based on age and smoking history. However, the individual benefits of screening are highly variable and potentially attenuated by co-morbidities such as advanced airflow limitation (AL).

Objective To examine the relationship between the presence and severity of AL and screening outcomes.

Methods This was a secondary analysis of 18 463 high-risk smokers, a substudy from the National Lung Screening Trial, who underwent pre-bronchodilator spirometry at baseline and median follow-up of 6.1 years. We used descriptive statistics and a competing risk proportional hazards model to examine differences in screening outcomes by chronic obstructive pulmonary disease severity group.

Results The risk of developing LC increased with worsening AL (effect size=0.34, p<0.0001), as did the risk of dying of LC (effect size=0.35, p<0.0001). While those with severe AL (Global Initiative for Obstructive Lung Disease, GOLD grade 3–4) had the highest risk of LC and the highest LC mortality, they also had fewer adenocarcinomas (effect size=−0.20, p=0.008) and a lower surgery rate (effect size=−0.16, p=0.014) despite comparable staging, and greater non-LC mortality relative to LC mortality (effect size=0.30, p<0.0001). In participants with no AL, screening with CT was associated with a significant reduction in LC deaths relative to chest X-ray (30.3%, 95% CI 4.5% to 49.2%, p<0.05). The clinically relevant but attenuated reduction in those with AL (18.5%, 95% CI −8.4% to 38.7%, p>0.05) could be attributed to GOLD 3–4, where no appreciable mortality reduction was observed.

Conclusion Despite a greater risk of LC, severe AL was not associated with any apparent reduction in LC mortality following screening.

  • Lung Cancer
  • COPD epidemiology

Data availability statement

All data relevant to the study are included in the article or uploaded as online supplemental information.

https://doi.org/10.1136/thorax-2022-219334

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    WHAT IS ALREADY KNOWN ON THIS TOPIC

    • Currently recommended eligibility criteria for lung cancer (LC) screening target those at greatest risk based on age and smoking history criteria. While this helps identify those at greatest risk of LC, it also identifies those most likely to have chronic obstructive pulmonary disease (COPD). While it is accepted that worsening airflow limitation (AL) (or COPD) confers a greater risk of LC, the question arises ‘Are the benefits from screening those with severe AL comparable to those with mild-to-moderate AL or normal lung function?’

    WHAT THIS STUDY ADDS

    • In this secondary analysis of 18 643 high-risk smokers from the National Lung Screening Trial, we found that although those with severe or very severe AL (Global Initiative for Obstructive Lung Disease, GOLD grade 3–4) have the highest risk for LC they also have lower surgical rates (despite comparable staging), more aggressive histology and higher rates of non-LC deaths. We suggest that these factors may contribute to an absence of any apparent reduction in LC mortality in this group following screening (‘poor responders’) and that their exclusion appears to improve screening efficiency.

    HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY

    • Current strategies to optimise LC screening focus primarily on increasing screening efficiency through improved risk prediction. However, these risk-based approaches also enrich for some comorbid diseases including severe COPD. While this group has the greatest risk of LC, this study shows they develop LC of a more aggressive histology, are less likely to undergo surgery and are more likely to die of non-LC causes (competing cause of death). The results of this study suggest the risk–benefit of screening for LC may be marginal for those with severe AL (GOLD 3–4) despite being at greatest risk (ie, is not linear). It also suggests that spirometric assessment may help improve screening efficiency by identifying those for whom the benefits of screening may be outweighed by the harms.

    Introduction

    Following the findings of three randomised controlled trials,1–3 annual CT screening for lung cancer (LC) is now more widely recommended in both the USA and Europe.4 5 While these studies reported relative risk reductions in LC specific mortality of between 20% and 33%, reduction in all-cause mortality was lower (0%–17% range).1–3 One potential explanation for this observation is the diluting effect of ‘competing cause of death’ on reducing overall mortality.6 In other words, while low-dose CT screening reduces deaths from LC, the benefit for all-cause mortality is attenuated by mortality from other smoking-related deaths, notably cardiorespiratory disease.7–10 Competing cause of death has been defined as ‘a failure to achieve improved life expectancy by preventing death from one disease due to death from another cause.’11 This concept is particularly relevant for LC because, relative to other screening populations, LC screening involves older heavy smokers for whom overall background morbidity and mortality is higher.6 This is due to coexisting smoking-related diseases, primarily chronic obstructive pulmonary disease (COPD) and cardiovascular disease.6–10 The impact of comorbid disease and premature death on LC screening outcomes is now the subject of considerable interest.9 12–14 because the benefit of screening may be diluted.15

    Studies have previously shown that airflow limitation (spirometric defined COPD), affects between 30% and 60% of those enrolled for LC screening.16–19 Airflow limitation is a marker for premature death from all causes20 and found to be unrecognised in 35%–70% of screening participants when spirometry is routinely performed.16–19 Although worsening airflow limitation increases the risk of developing LC in the National Lung Screening Trial (NLST),21 in a preliminary analysis, we observed that it was also associated with an almost halving in lung-cancer specific mortality relative to those with normal lung function.22 We propose that the increased risk of LC associated with worsening airflow limitation is also associated with a greater risk of dying from a cause other than LC.6 12 We and others have also observed that LC in smokers with airflow limitation may be more aggressive with more squamous cell and less adenocarcinoma subtypes (histology shift).16 23 This raises two questions, ‘As airflow limitation worsens, is there a differential effect on LC-specific mortality relative to other causes of death?’ and ‘How might the presence of severe airflow limitation attenuate the benefits of CT screening?’

    In this secondary analysis of the American College of Radiology Imaging Network (ACRIN) subcohort of the NLST participants (N=18 463), where baseline spirometry was available and the risk of LC could be estimated, we undertook this study to examine the relationship between the presence and severity of airflow limitation and outcomes from screening.

    Methods

    Subjects

    The recruitment and study design of the full NLST, involving 53 452 screening participants has been described elsewhere.1 In the ACRIN subcohort of the NLST, which included participants from 23 screening centres (N=18 840), demographic data were collected through an extensive questionnaire and prebronchodilator pulmonary function tests recorded at baseline (online supplementary methods and supplementary figure 1).

    Supplemental material

    Clinical and demographic variables

    Demographic variables and clinical variables outlining the subject characteristics at baseline (N=18 463), and prospectively diagnosed LC characteristics (N=785), are described in detail in online supplementary methods.

    Screening outcomes

    • LC cases: included those diagnosed during the trial (N=757) or during postmortem examination (N=28).

    • Stage shift: the proportion of patients diagnosed with LC in stage 1 or 2 was determined for each screening group.

    • LC surgery: the proportion of LC cases that underwent surgery.

    • Mortality: LC and non-LC deaths during follow-up as ascertained through review of clinical records and death certification (total=1372).

    Statistical analysis

    We first determined summary statistics for clinical and demographic variables and study outcomes by COPD severity groups, and tested for overall associations by using χ2, Fisher’s exact or analysis of variance tests, as appropriate. We also reported effect sizes as Goodman and Kruskal’s gamma statistic, a measure of rank correlation with a range of (−1 to 1); values closer to 0 indicate lower association between compared variables (see online supplementary section). We next compared absolute differences among LC cases between screening groups (CT vs chest X-ray (CXR)) within each COPD severity group for: (1) stage 1–2 diagnoses, (2) adenocarcinoma histology, (3) surgical treatment following diagnosis and (4) LC death. These absolute differences were expressed as percentages with 95% CIs. We then calculated several LC death statistics and their 95% CI from the full screening population by comparing screening groups within each COPD severity level (see online supplementary methods). In further online supplementary analyses, we conducted competing risk proportional hazards analyses (Fine and Gray subdistributional models—see Supplementary Methods). All analyses were performed using SAS (V.9.4, SAS Institute).

    Results

    Baseline comparison of demographic and clinical variables

    From the total cohort of 18 643 NLST subjects, there were 12 303 controls with no airflow limitation (66.6%), 1499 had Global Initiative for Obstructive Lung Disease (GOLD) 1 (8.1%), 3412 had GOLD 2 (18.5%) and 1249 had GOLD 3–4 (6.8%), airflow limitation (table 1). Airflow limitation was associated with the following differences: older age, being male, greater duration of smoking, greater pack years and greater rate of current smoking. Worsening airflow limitation was associated with modestly reduced body mass index. GOLD 3–4 disease was associated with the greatest cigarettes per day, pack years, history of COPD; lowest educational level, worst lung function and the most respiratory comorbid disease. GOLD 3–4 was also associated with the most heart disease. Only 56% of those with GOLD 3–4 disease reported a prior diagnosis of COPD compared with 18% and 30% in those with GOLD 1 and 2, respectively (table 1). Airflow limitation was not associated with ethnicity, or family history of LC.

    View this table:
    Table 1

    Baseline demographic variables according to the presence and severity of airflow limitation

    LC and mortality outcomes

    From the total cohort, the risk of developing LC increased with worsening airflow limitation (effect size=0.34, p<0.0001) (table 2, figure 1). Similarly, the risk of dying of LC also increased according to worsening airflow limitation (effect size=0.35, p<0.0001). GOLD 3–4 patients had the highest rates of LC diagnosis and LC mortality. With increasing airflow limitation, there was a decreasing prevalence of adenocarcinoma (effect size=−0.20, p=0.008) and less surgery (effect size=−0.16, p=0.014), despite comparable staging. Compared with controls, the GOLD 3–4 COPD group was associated with a significantly greater prevalence of non-small cell lung carcinoma–not otherwise specified (NSCLC-NOS) histology and less adenocarcinoma (table 2 and online supplemental table 1). There was no effect on stage shift by COPD severity. GOLD 3–4 COPD was associated with a lower prevalence of LCs in the prevalent (baseline, T0) scan, less screen-detected LCs and greater LC prevalence during the follow-up (non-screening) interval but these differences were not statistically different. With increasing airflow limitation, the non-LC mortality increased at a greater rate than for LC mortality (figure 1). This divergence was attributed in the main to increasing cardiorespiratory deaths in the GOLD 3–4 group.

    Figure 1
    Figure 1

    Lung cancer prevalence and cause-specific mortality per 100 screened according to gold grade (data from table 2). The increase in non-lung cancer deaths (red) is steeper than for lung cancer deaths (orange) hence divergence in overall mortality across worsening airflow limitation. Worsening airflow limitation is associated with increases in cause-specific mortality; notably cardiovascular (black) and respiratory deaths (blue) were greatest in those with GOLD 3–4. GOLD, Global Initiative for Obstructive Lung Disease.

    View this table:
    Table 2

    Characteristics of LC cases and non-LC mortality stratified by the presence and severity of airflow limitation

    Screening outcomes by COPD group

    Table 3 includes screening outcomes among LC cases and among the full population by screening group and COPD severity. Group sizes used to support these calculations are also included. Among LC cases, outcomes include absolute changes in stage 1–2 diagnoses, adenocarcinoma histology, surgical treatment following diagnosis and LC death. Among the full population we describe both relative and absolute differences in mortality from LC (see Supplementary Methods).

    View this table:
    Table 3

    Outcomes following randomisation to the CT and CXR arms according to the presence and/or severity of airflow limitation (GOLD grade)

    For the controls (no airflow limitation), randomisation to the CT arm favoured stage shift to early-stage cancers and significant reductions in LC deaths in relative terms (30.3% reduction, 95% CI 4.5% to 49.2%) and absolute terms (4.6 LC deaths averted per 1000 screened, 95% CI 0.4 to 8.4). For those with airflow limitation (GOLD grades 1–4), there was an attenuated benefit in those randomised to CT with a non-significant reduction in LC deaths (18.5% relative reduction, 95% CI −8.4% to 38.7%) relative to CXR. It is notable that screening benefits due to stage shift and LC mortality reduction were reduced as airflow limitation increased. For those with severe or very severe airflow limitation (GOLD 3–4), there was no apparent stage shift, adenocarcinoma histology shift, or reduction in LC mortality; further, there was a negative estimate for number needed to screen (NNS). This contrasts with those with GOLD one airflow limitation with a comparable sample size (and similar powering); figure 2A further illustrates this contrast. We also note that despite lower rates of LC-related surgery with worsening airflow limitation (table 2), there was consistently greater surgery in those randomised to CT relative to CXR in all groups (including GOLD grade 3–4, 47% vs 33%) (table 3). Interestingly, while 47% of LC cases in the CT arm underwent surgery for GOLD 3–4 (table 3, figure 2), there was no meaningful stage shift and no apparent reduction in LC mortality. If this group (GOLD 3–4) were excluded from screening on the basis that harms may outweigh the benefits (figure 2B), we found screening efficiency marginally increased as indicated by a greater relative reduction in LC mortality: 29.0% (95% CI 10.6% to 43.7%) up from 24.9% in the full group (95% CI 7.1% to 39.2%), and a reduced NNS to avert one LC death: 174, 95% CI (57 to 290) down from 190, 95% CI (50 to 329) in the full group.

    Figure 2
    Figure 2

    Screening outcomes favouring randomisation to the CT arm relative to the CXR arm in the NLST according to gold grade (data from table 3). (A) Subgrouped by severity of airflow limitation (B) grouping GOLD 1 and 2 (based on comparable outcomes on lung cancer mortality—see table 3). GOLD groups 1 and 2 have been combined on the basis the lung cancer mortality benefit favouring CT are comparable; ORs of 0.75 (95% CI 0.41 to 1.39) and 0.72 (95% CI 0.47 to 1.10) and absolute lung cancer deaths averted per 1000 of 7.8 (95% CI −10.4 to 23.3) and 8.3 (95% CI −3.1 to 18.5) for gold groups 1 and 2, respectively. CXR, chest X-ray; GOLD, Global Initiative for Obstructive Lung Disease; NLST, National Lung Screening Trial.

    Factors contributing to LC deaths

    We found similar results from a competing risk proportional hazards model (Fine and Gray) for LC death, adjusted for important clinical and demographic predictors (Supplementary Methods and table 2). Although the interaction between screening arm and COPD severity was not significant in the model (p=0.53), the stratified results indicated a trend towards CT screening advantage for the non-COPD group (HR 0.76 95% CI 0.56 to 1.03, p=0.073) and the GOLD 1–2 group (HR 0.73, 95% CI 0.52 to 1.02, p=0.063), but statistically non-significant results for the GOLD 3–4 group (online supplemental table 2). We conclude from the model that age, pack years, years since quitting, history of self-reported COPD, emphysema, asthma and diabetes all contribute to dying from LC. The cumulative incidence function plots (online supplemental figure 2) further illustrate the increasing and diverging risk for LC death as COPD severity increases, with a corresponding decline in screening benefit.

    Discussion

    In this analysis of the ACRIN arm of the NLST, including 18 463 high-risk subjects, we have examined the effect of underlying airflow limitation on outcomes from LC screening. Although airflow limitation is associated with an increased risk of LC,22 in this study having severe airflow limitation (GOLD 3–4) was associated with no apparent benefit from CT-based LC screening. Specifically, the reduced improvement in LC cancer mortality we report in the total group with airflow limitation, 18% vs 30% in those with no airflow limitation, could be attributed almost entirely to those with the most severe airflow obstruction (GOLD 3–4 grade) and greatest respiratory comorbidity. While GOLD 3–4 subjects represented nearly 7% of the entire ACRIN cohort, they accounted for 14.6% of LCs. We note that for GOLD 3–4 subjects, 47% of the LC identified in the CT arm underwent surgery although no benefit from screening and treatment was observed. The basis of this finding likely stems from one or a combination of factors related to the screening subject or their LC.6 Despite GOLD 3–4 screening participants having greater respiratory-related comorbid disease and greater cardiorespiratory deaths, only 56% were aware they had COPD. This group also developed LCs that were less likely to be detected by screening, were of a more aggressive histology (more squamous cell and NSCLC-NOS but less adenocarcinomas),16 23 24 and experienced lower surgical rates for their cancers. These LC characteristics likely underpin the lack of stage shift in the CT arm relative to CXR in this group. Collectively, these findings demonstrate that among those with GOLD 3–4 airflow limitation, and at greatest risk of LC, nearly half underwent work-up and surgery yet no apparent benefit from screening was observed. In the clinical setting where LC screening is targeted to those who stand to gain the most from screening,19 spirometry and respiratory comorbidity may help identify those for whom screening may expose them to greater harm than benefit.

    One important implication of this finding is that the relationship between risk and benefit from screening is not linear.6 12 Specifically, those at greatest risk of LC according to their spirometry,21 actually do not benefit most from screening. In fact, this study suggests that spirometric assessment of smokers eligible for screening will identify those with GOLD 3–4 airflow limitation who appear to be ‘poor responders’ to LC screening and for whom screening may be more harmful than beneficial. A second implication of this study is that including spirometry during screening may provide very useful information for the screening participant and their physician by identifying undiagnosed COPD or quantifying severity of airflow limitation.6 16–19 A third and more important implication of this finding is that to optimise LC screening benefits (and efficiency) it might be better to focus on those with the best outcomes rather than focus solely on those at greatest risk of LC. LC screening is quite different to other cancer screening programmes because those eligible for screening are enriched to have the greatest risk but will include many with a shortened life expectancy.6–12 As pulmonary function tests are also closely linked to life expectancy,20 better reflecting biological age rather than chronological age, their routine use in LC screening may help identify who derives the least benefit from screening. For these reasons, we propose that quantifying airflow limitation provides useful information about the outcomes (responsiveness) and risks for smokers undergoing LC screening.6 13 14

    There are several factors that might contribute to the poor outcomes in NLST subjects with GOLD 3–4 disease. Consistent with studies in non-screened LC, we have shown airflow limitation in screened subjects was associated with more aggressive types of LC, specifically squamous cell and NSCLC-NOS subtypes.16 23–25 We and others have linked pre-existing airflow limitation in LC subjects with shorter volume doubling times.26 27 This may mean that the most aggressive LCs are less amenable to detection and successful treatment through screening. Our results support this by showing those with GOLD 3–4 had fewer screen-detected cancers and less surgery overall (independent of screening arm). These differences may explain the lack of stage shift in this group when comparing CT with CXR and certainly also explains the lack of benefit in reducing LC mortality. Another possible explanation for there being no benefit from screening in GOLD 3–4 subjects is that they experienced much higher non-LC deaths relative to their LC death rate (divergence in figure 1). This could be attributed to the higher rates of pre-existing comorbid respiratory disease and high rates of cardiorespiratory death during screening.6 25 When life expectancy is factored into assessing the benefits of LC screening, it appears the best outcomes are achieved when those with intermediate risk are targeted.6 12 28 This combines increased risk of LC with greater relative reduction in LC deaths and greater long-term survival.

    A key result from this study is that of the 49 LC deaths averted by using CT-based screening in the whole cohort, 27 deaths were averted in those with normal lung function and 23 deaths were averted in those with GOLD 1–2 disease (table 3). There was one excess LC death in the CT arm over the CXR arm for those with GOLD 3–4 disease. Thus, in our secondary analysis, we found no benefit in screening for LC in the latter group. In fact, screening was more efficient (lower NNS) when this group of poor responders was excluded. We note the prevalence of GOLD 3–4 airflow limitation was 7% in the NLST-ACRIN study and between 4% and 8% in other screening studies.17–19 Given there was likely to be greater harm from the work up and treatment of LCs in those with severe COPD,29 we suggest the net harm may outweigh the benefit. This argues strongly for identifying eligible smokers with GOLD 3–4 disease and greater consideration of whether this group should be offered surgery following screening. While stereotactic-based radiotherapy has been shown in observational studies to achieve comparable short-term survival to those receiving surgery in unscreened LC patients with GOLD 3–4 COPD,30 the long-term benefits relative to complications from screening and investigating this group remains less clear.8 Prospective studies comparing mortality reduction according to lung function are needed to confirm our findings. We suggest that as the risk of LC increases, the potential for doing harm may also increase when selection criteria for screening targets only those at greatest risk. This demonstrates the greater utility of an outcomes-based approach over a risk-based approach to screening. In an outcomes-based approach, smokers eligible for screening based on age and pack years criteria are reassessed with regards to their ‘responsiveness’ to LC screening according to comorbid diseases like severe COPD where the impact on screening outcomes, and thus the benefit to harm ratio, are significantly altered.

    There are several strengths and weaknesses to this study. This subanalysis included data for over 18 000 screening subjects from 23 different sites who underwent baseline spirometry and followed for a median of 6.1 years. Despite this large study size, the number of LCs diagnosed during this study was only 785 and this significantly limited our ability to determine what variables contributed most to the poor outcomes we report for the GOLD 3–4 group. This means that after stratification by GOLD grade, our analyses were underpowered. That said, the primary clinical end point of LC death allowed us to examine differences in outcome by screening. Other weakness includes no data on biopsy rates, procedural complications from nodule work up or perioperative care. This may be important as others have shown that subjects with COPD have more nodules to follow-up, greater complications during nodule workup and more complications from surgery.29 Lastly, we note the NLST cohort may not best represent those undergoing screening for LC in community-based studies and that ongoing prospective studies comparing outcomes are required to confirm our findings.

    In conclusion, this study demonstrates that increasing risk of LC does not necessarily translate into increasing benefit from screening in a simple linear relationship. More importantly the findings show routine use of spirometry helps identify those with severe airflow limitation conferring a reduced life expectancy, greater risk for aggressive LC and greater mortality risk from non-LC causes. These observations suggest that routine use of spirometry may help identify this largely unrecognised but important ‘poor responder’ subgroup for whom LC screening may cause more harm than benefit.

    Data availability statement

    All data relevant to the study are included in the article or uploaded as online supplemental information.

    Ethics statements

    Patient consent for publication

    References

    1. ↵
      1. National Lung Screening Trial Research Team,
      2. Aberle DR ,
      3. Adams AM , et al
      . Reduced lung-cancer mortality with low-dose computed tomographic screening. N Engl J Med 2011;365:395–409.doi:10.1056/NEJMoa1102873 pmid:http://www.ncbi.nlm.nih.gov/pubmed/21714641
      OpenUrlCrossRefPubMedWeb of Science
    2. ↵
      2. Paci E ,
      3. Puliti D ,
      4. Lopes Pegna A , et al
      . Mortality, survival and incidence rates in the ITALUNG randomised lung cancer screening trial. Thorax 2017;72:825–31.doi:10.1136/thoraxjnl-2016-209825 pmid:http://www.ncbi.nlm.nih.gov/pubmed/28377492
      OpenUrlAbstract/FREE Full Text
    3. ↵
      2. de Koning HJ ,
      3. van der Aalst CM ,
      4. de Jong PA , et al
      . Reduced lung-cancer mortality with volume CT screening in a randomized trial. N Engl J Med 2020;382:503–13.doi:10.1056/NEJMoa1911793 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31995683
      OpenUrlCrossRefPubMed
    4. ↵
      2. Ito Fukunaga M ,
      3. Wiener RS ,
      4. Slatore CG
      . The 2021 us preventive services Task force recommendation on lung cancer screening: the more things stay the Same…. JAMA Oncol 2021;7:684–6.doi:10.1001/jamaoncol.2020.8376 pmid:http://www.ncbi.nlm.nih.gov/pubmed/33687430
      OpenUrlPubMed
    5. ↵
      2. van der Aalst CM ,
      3. Ten Haaf K ,
      4. de Koning HJ
      . Implementation of lung cancer screening: what are the main issues? Transl Lung Cancer Res 2021;10:1050–63.doi:10.21037/tlcr-20-985 pmid:http://www.ncbi.nlm.nih.gov/pubmed/33718044
      OpenUrlPubMed
    6. ↵
      2. Young RP ,
      3. Hopkins RJ
      . Chronic obstructive pulmonary disease (COPD) and lung cancer screening. Transl Lung Cancer Res 2018;7:347–60.doi:10.21037/tlcr.2018.05.04 pmid:http://www.ncbi.nlm.nih.gov/pubmed/30050772
      OpenUrlPubMed
    7. ↵
      2. Tanner NT ,
      3. Dai L ,
      4. Bade BC , et al
      . Assessing the generalizability of the National lung screening trial: comparison of patients with stage 1 disease. Am J Respir Crit Care Med 2017;196:602–8.doi:10.1164/rccm.201705-0914OC pmid:http://www.ncbi.nlm.nih.gov/pubmed/28722466
      OpenUrlPubMed
    8. ↵
      2. Iaccarino JM ,
      3. Steiling KA ,
      4. Wiener RS
      . Lung cancer screening in a safety-net Hospital: implications of screening a real-world population versus the National lung screening trial. Ann Am Thorac Soc 2018;15:1493–5.doi:10.1513/AnnalsATS.201806-389RL pmid:http://www.ncbi.nlm.nih.gov/pubmed/30199643
      OpenUrlPubMed
    9. ↵
      2. Callister MEJ ,
      3. Sasieni P ,
      4. Robbins HA
      . Overdiagnosis in lung cancer screening. Lancet Respir Med 2021;9:7–9.doi:10.1016/S2213-2600(20)30553-1 pmid:http://www.ncbi.nlm.nih.gov/pubmed/33412118
      OpenUrlPubMed
    10. ↵
      2. Esserman LJ ,
      3. Thompson IM ,
      4. Reid B
      . Overdiagnosis and overtreatment in cancer: an opportunity for improvement. JAMA 2013;310:797–8.doi:10.1001/jama.2013.108415 pmid:http://www.ncbi.nlm.nih.gov/pubmed/23896967
      OpenUrlCrossRefPubMedWeb of Science
    11. ↵
    12. ↵
      2. Rivera MP ,
      3. Tanner NT ,
      4. Silvestri GA , et al
      . Incorporating coexisting chronic illness into decisions about patient selection for lung cancer screening. An official American thoracic Society research statement. Am J Respir Crit Care Med 2018;198:e3–13.doi:10.1164/rccm.201805-0986ST pmid:http://www.ncbi.nlm.nih.gov/pubmed/30004250
      OpenUrlCrossRefPubMed
    13. ↵
      2. Silvestri GA ,
      3. Young RP
      . Strange bedfellows: the interaction between COPD and lung cancer in the context of lung cancer screening. Ann Am Thorac Soc 2020;17:810–2.doi:10.1513/AnnalsATS.202005-433ED pmid:http://www.ncbi.nlm.nih.gov/pubmed/32609030
      OpenUrlPubMed
    14. ↵
      2. Young RP ,
      3. Hopkins RJ ,
      4. Gamble GD , et al
      . Incorporating baseline lung function in lung cancer screening. Chest 2021;159:1664–9.doi:10.1016/j.chest.2020.10.070
      OpenUrl
    15. ↵
      2. Cheung LC ,
      3. Berg CD ,
      4. Castle PE , et al
      . Life-gained-based versus risk-based selection of smokers for lung cancer screening. Ann Intern Med 2019;171:623–32.doi:10.7326/M19-1263 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31634914
      OpenUrlPubMed
    16. ↵
      2. Young RP ,
      3. Duan F ,
      4. Chiles C , et al
      . Airflow limitation and histology shift in the National lung screening trial. The NLST-ACRIN cohort substudy. Am J Respir Crit Care Med 2015;192:1060–7.doi:10.1164/rccm.201505-0894OC pmid:http://www.ncbi.nlm.nih.gov/pubmed/26199983
      OpenUrlPubMed
    17. ↵
      2. Balata H ,
      3. Harvey J ,
      4. Barber PV , et al
      . Spirometry performed as part of the Manchester community-based lung cancer screening programme detects a high prevalence of airflow obstruction in individuals without a prior diagnosis of COPD. Thorax 2020;75:655–60.doi:10.1136/thoraxjnl-2019-213584 pmid:http://www.ncbi.nlm.nih.gov/pubmed/32444437
      OpenUrlAbstract/FREE Full Text
    18. ↵
      2. Ruparel M ,
      3. Quaife SL ,
      4. Dickson JL , et al
      . Prevalence, symptom burden, and underdiagnosis of chronic obstructive pulmonary disease in a lung cancer screening cohort. Ann Am Thorac Soc 2020;17:869–78.doi:10.1513/AnnalsATS.201911-857OC pmid:http://www.ncbi.nlm.nih.gov/pubmed/32164439
      OpenUrlPubMed
    19. ↵
      2. Goffin JR ,
      3. Pond GR ,
      4. Puksa S , et al
      . Chronic obstructive pulmonary disease prevalence and prediction in a high-risk lung cancer screening population. BMC Pulm Med 2020;20:300.doi:10.1186/s12890-020-01344-y pmid:http://www.ncbi.nlm.nih.gov/pubmed/33198781
      OpenUrlPubMed
    20. ↵
      2. Shavelle RM ,
      3. Paculdo DR ,
      4. Kush SJ
      . Life expectancy and years of life lost in chronic obstruvctive pulmonary disease: findings from the NHANES III follow-up study. Int J COPD 2019;4:137–48.
      OpenUrl
    21. ↵
      2. Hopkins RJ ,
      3. Duan F ,
      4. Chiles C , et al
      . Reduced expiratory flow rate among heavy smokers increases lung cancer risk. results from the National lung screening Trial-American College of radiology imaging network cohort. Ann Am Thorac Soc 2017;14:392–402.doi:10.1513/AnnalsATS.201609-741OC pmid:http://www.ncbi.nlm.nih.gov/pubmed/28076701
      OpenUrlPubMed
    22. ↵
      2. Young RP ,
      3. Duan F ,
      4. Greco E
      . Lung cancer-specific mortality reduction with CT screening: outcomes according to airflow limitation in the ACRIN-NLST study (N=18,475). Am J Respir Crit Care Med 2016;193:A6166.
    23. ↵
      2. Papi A ,
      3. Casoni G ,
      4. Caramori G , et al
      . Copd increases the risk of squamous histological subtype in smokers who develop non-small cell lung carcinoma. Thorax 2004;59:679–81.doi:10.1136/thx.2003.018291 pmid:http://www.ncbi.nlm.nih.gov/pubmed/15282388
      OpenUrlAbstract/FREE Full Text
    24. ↵
      2. Righi L ,
      3. Vavalà T ,
      4. Rapa I , et al
      . Impact of non-small-cell lung cancer-not otherwise specified immunophenotyping on treatment outcome. J Thorac Oncol 2014;9:1540–6.doi:10.1097/JTO.0000000000000271 pmid:http://www.ncbi.nlm.nih.gov/pubmed/25521399
      OpenUrlPubMed
    25. ↵
      2. Kumar V ,
      3. Cohen JT ,
      4. van Klaveren D , et al
      . Risk-Targeted lung cancer screening: a cost-effectiveness analysis. Ann Intern Med 2018;168:161–9.doi:10.7326/M17-1401 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29297005
      OpenUrlCrossRefPubMed
    26. ↵
      2. Young RP ,
      3. Hopkins RJ
      . Estimating overdiagnosis of lung cancer. Ann Intern Med 2013;158:635.doi:10.7326/0003-4819-158-8-201304160-00013 pmid:http://www.ncbi.nlm.nih.gov/pubmed/23588755
      OpenUrlPubMed
    27. ↵
      2. Maisonneuve P ,
      3. Veronesi G ,
      4. Bertolotti R
      . Estimating overdiagnosis of lung cancer--reply. Ann Intern Med 2013;158:635–6.doi:10.7326/0003-4819-158-8-201304160-00014 pmid:http://www.ncbi.nlm.nih.gov/pubmed/23588756
      OpenUrlPubMed
    28. ↵
      2. Ten Haaf K ,
      3. van der Aalst CM ,
      4. de Koning HJ
      . Clinically detected non-aggressive lung cancers: implications for overdiagnosis and overtreatment in lung cancer screening. Thorax 2018;73:407–8.doi:10.1136/thoraxjnl-2017-211149 pmid:http://www.ncbi.nlm.nih.gov/pubmed/29440512
      OpenUrlFREE Full Text
    29. ↵
      2. Iaccarino JM ,
      3. Silvestri GA ,
      4. Wiener RS
      . Patient-Level trajectories and outcomes after low-dose CT screening in the National lung screening trial. Chest 2019;156:965–71.doi:10.1016/j.chest.2019.06.016 pmid:http://www.ncbi.nlm.nih.gov/pubmed/31283920
      OpenUrlPubMed
    30. ↵
      2. Palma D ,
      3. Lagerwaard F ,
      4. Rodrigues G , et al
      . Curative treatment of stage I non-small-cell lung cancer in patients with severe COPD: stereotactic radiotherapy outcomes and systematic review. Int J Radiat Oncol Biol Phys 2012;82:1149–56.doi:10.1016/j.ijrobp.2011.03.005
      OpenUrlCrossRefPubMed

    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 RPY and RJS contributed to the conception and design; acquisition, analysis and interpretation; drafting and review for important intellectual content and final approval of the manuscript. GS contributed to the analysis and interpretation; drafting and review for important intellectual content and final approval of the manuscript. RCW and GDG contributed to bio statistical analysis, drafting, review and final approval of the manuscript.

    • 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.

    Linked Articles