Policy Review

European position statement on lung cancer screening

Matthijs Oudkerk, Anand Devaraj, Rozemarijn Vliegenthart, Thomas Henzler, Helmut Prosch, Claus P Heussel, Gorka Bastarrika, Nicola Sverzellati,
Mario Mascalchi, Stefan Delorme, David R Baldwin, Matthew E Callister, Nikolaus Becker, Marjolein A Heuvelmans, Witold Rzyman,
Maurizio V Infante, Ugo Pastorino, Jesper H Pedersen, Eugenio Paci, Stephen W Duffy, Harry de Koning, John K Field

Lung cancer screening with low-dose CT can save lives. This European Union (EU) position statement presents the Lancet Oncol 2017
available evidence and the major issues that need to be addressed to ensure the successful implementation of low-dose Published Online
CT lung cancer screening in Europe. This statement identified specific actions required by the European lung cancer November 27, 2017
http://dx.doi.org/10.1016/
screening community to adopt before the implementation of low-dose CT lung cancer screening. This position
S1470-2045(17)30861-6
statement recommends the following actions: a risk stratification approach should be used for future lung cancer low-
Center for Medical Imaging,
dose CT programmes; that individuals who enter screening programmes should be provided with information on the University Medical Center
benefits and harms of screening, and smoking cessation should be offered to all current smokers; that management of Groningen, University of
detected solid nodules should use semi-automatically measured volume and volume-doubling time; that national Groningen, Groningen,
Netherlands
quality assurance boards should be set up to oversee technical standards; that a lung nodule management pathway
(Prof M Oudkerk MD,
should be established and incorporated into clinical practice with a tailored screening approach; that non-calcified R Vliegenthart MD,
baseline lung nodules greater than 300 mm³, and new lung nodules greater than 200 mm³, should be managed in M A Heuvelmans MD);
multidisciplinary teams according to this EU position statement recommendations to ensure that patients receive the Department of Radiology,
Royal Brompton Hospital,
most appropriate treatment; and planning for implementation of low-dose CT screening should start throughout
London, UK (A Devaraj MD);
Europe as soon as possible. European countries need to set a timeline for implementing lung cancer screening. Institute of Clinical Radiology
and Nuclear Medicine,
Introduction comprehensive literature search for papers on lung University Medical Centre
Mannheim, Medical Faculty
Lung cancer screening with low-dose CT can save lives, cancer screening and, through in-depth discussions,
Mannheim, Heidelberg
and this method will probably be embraced by national developed this EU position statement consensus. University, Mannheim,
health organisations throughout Europe in the future. The structure of this document not only reflects the Germany (T Henzler MD);
The results from the US National Lung Cancer Screening available evidence that addresses the major questions Department of Biomedical
Imaging and Image-guided
Trial (NLST)1 on reduced lung cancer mortality and from concerning the delivery of a successful screening Therapy, Medical University of
seven pilot trials­2–8 within Europe on other aspects of low- intervention, but also highlights any issues that still Vienna, Vienna General
dose CT screening have provided sufficient evidence for need  to be resolved for successful implementation. Hospital, Vienna, Austria
Europe to start planning for lung cancer screening while Contributions to this EU position statement were (H Prosch MD); Translational
Research Unit and Department
mortality data from the NELSON trial2 are awaited. provided by a team of clinicians and scientists expert in of Diagnostic and
This European Union (EU) position statement CT as the method of choice for lung cancer screening. Interventional Radiology with
describes the current status of lung cancer screening The requirement for an EU position statement stems Nuclear Medicine, Thoraxklinik,
and sets out the essential elements needed to ensure from the need to provide European recommendations on Heidelberg University,
Heidelberg, Germany
the development of effective European screening CT screening that will assist the EU Commission and (C P Heussel MD); Translational
programmes. The EU position statement expert group national health agencies in beginning to plan the Lung Research Centre
comprises individuals from eight European countries implementation of lung cancer screening within the next Heidelberg, German Centre for
who have been actively engaged in the planning and 2 years, and to avoid opportunistic and uncontrolled Lung Research, Heidelberg,
Germany (C P Heussel);
execution of randomised controlled screening trials in screening. Moreover, since the publication of the NLST Department of Diagnostic and
Europe,9 who are involved in the clinical management of results in 2011,1 an EU position statement on the value of Interventional Radiology,
patients with lung cancer and lung nodules, and who CT screening for lung cancer is now a crucial necessity. University-Hospital
have developed relevant clinical practice guidelines on The focus of this EU position statement is restricted to Heidelberg, Heidelberg,
Germany (C P Heussel);
smoking cessation, recruitment of high risk participants, lung cancer screening with low-dose CT and the early Department of Diagnostic and
patient information literature, as well as CT screening detection of lung nodules before clinical work-up, and Interventional Radiology
protocols, CT scan radiology reporting, and the clinical does not address the entirety of work-up and treatment Department of Radiology,
Clínica Universidad de Navarra,
management of CT-detected nodules. These experts choices. Since new randomised controlled trials of low-
Pamplona, Spain
represent all the specialties and professions involved in dose CT screening that are powered to allow conclusions (G Bastarrika MD); Radiology,
delivering successful lung cancer screening programmes about mortality reduction are highly unlikely, our Department of Medicine and
in Europe. The emphasis of this EU position statement recommendations are based on the current available data. Surgery, University of Parma,
Parma, Italy
focuses on the actual implementation of CT lung cancer Data provided by several studies2,6,8 are sufficient to make
(Prof N Sverzellati MD);
screening programmes in Europe by radiologists, recommendations concerning the minimisation of false Department of Clinical and
supported by epidemiologists, pulmonologists, and positive results in both screen-detected and non-screen Experimental Biomedicine,
thoracic surgeons, in the full context of clinical lung detected nodules. The need for non-contrast-enhanced University of Florence,
Florence, Italy
cancer diagnosis and treatment. These individuals low-dose interval imaging should not be considered a
(M Mascalchi MD); Division of
comprise the core membership of the EU Lung cancer false-positive test because the individual is not undergoing Radiology (Prof S Delorme MD)
CT Screening Implementation Group (EU-LSIG) and an invasive clinical work-up and therefore the risk of and Division of Cancer
have prepared this EU position statement. We did a physical harm is very low. Furthermore, evidence10,11 Epidemiology

www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6 1

 Policy Review
(Prof N Becker PhD), German
Cancer Research Center,
Heidelberg, Germany;
Respiratory Medicine Unit,
David Evans Research Centre,
Nottingham University
Hospitals, Nottingham, UK
(Prof D R Baldwin MD);
Department of Respiratory
Medicine, Leeds Teaching
Hospitals, St James’s University
Hospital, Leeds, UK
(M E Callister MD); Department
of Thoracic Surgery, Medical
University of Gdańsk, Gdańsk,
Poland (Prof W Rzyman MD);
Thoracic Surgery Department,
University and Hospital Trust—
Azienda Ospedaliera
Universitaria Integrata, Verona,
Italy (M V Infante MD);
Department of Thoracic
Surgery, Istituto Nazionale
Tumori, Milan, Italy
(U Pastorino MD); Department
of Cardiothoracic Surgery,
Rigshospitalet, University of
Copenhagen, Copenhagen,
Denmark (J H Pedersen MD);
ISPO Cancer Research and
Prevention Institute Tuscany
Region, Florence, Italy
(Prof E Paci MD); Wolfson
Institute of Preventive
Medicine, Barts and The
London School of Medicine and
Dentistry, London, UK
(Prof S W Duffy PhD);
Department of Public Health,
Erasmus MC, Rotterdam,
Netherlands
(Prof H J de Koning MD); and
Roy Castle Lung Cancer
Research Programme,
Department of Molecular and
Clinical Cancer Medicine,
University of Liverpool,
Liverpool, UK (Prof J K Field PhD)
Correspondence to:
Prof John K Field, Roy Castle Lung
Cancer Research Programme,
Department of Molecular and
Clinical Cancer Medicine,
University of Liverpool,
Liverpool L7 8TX, UK
j.k.field@liv.ac.uk
2 www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6
shows that the psychological distress caused is transient planned because the stop criteria of a 20% reduction in
and smoking cessation rates increase among those who lung cancer mortality rate with low-dose CT had been
require interval imaging. reached in a periodic planned interim analysis
This position statement represents a balance of the compared with that achieved by chest x-ray. The trial
available data and therefore reflects which approaches are also showed a 6·7% reduction in all-cause mortality
supported by good evidence, where further evidence is with low-dose CT screening (1877 deaths in the low-
needed to implement effective screening programmes, dose CT group compared with 2000 deaths in the
and where practical implications for lung cancer radiography group).1
screening can already be drawn from the available There is increasing evidence of the effectiveness of CT
knowledge. screening from several pilot trials in Europe and from
the NELSON trial publications2,16,17 (table). However, we
Diagnostic tests for lung cancer detection need to remain aware of the implications and problems
CT has progressed to be the best method for lung cancer associated with the work-up of suspicious nodules (ie,
screening. Previous lung cancer screening trials in the the invasiveness of biopsies, waiting time until a final
1980s using chest x-rays with and with­ out sputum decision).
cytology showed no significant survival advantage,12,13 The high false-positive rates both in the initial
which led to inactivity in this field of research for more screening and subsequent screening rounds, as
than two decades. The first publication on lung CT reported in the NLST, need to be reduced to ensure that
screening in 199914 ignited interest in this field again. harmful effects on participants are kept to a minimum.
Other diagnostic methods might have a future potential This reduction is best achieved by accurate interval
in lung cancer screening but no trials are yet available to imaging with the latest and most precise methods,
support clinical use.15 particularly semi-automated volumetric analysis rather
Earlier trials using CT as a screening method provided than manual measurement of maximum diameter, as
evidence not only on the effectiveness of lung cancer already implemented by several trials.2,7,8 Furthermore,
screening, but also on the natural history of the disease. the definition of false positives also has a major bearing
The debate continued on the ability of CT screening to on how we interpret false-positive data. The NELSON,16
reduce mortality until the NLST was published,1 in MILD,3 and UKLS8 trials defined false positives using
which 53  454 patients were randomly assigned to baseline data as patients who required a referral to the
receive either low-dose CT or a chest x-ray for screening. pulmonologist and who required a further diagnostic
This trial had its results reported 1 year earlier than investigation (3·5%), but who subsequently did not
Recruitment Recruitment criteria Screening methods
period
Randomised controlled trials
NLST1 2002–04 Age 55–75 years, ≥30 PY smoker, quit smoking <15 years earlier Annual low-dose CT vs chest x-ray for 3 years
MILD3 2005–11 Age >49 years, ≥20 PY smoker, quit smoking <10 years earlier, Three groups: no screen, annual screen, and biennial
no cancers within past 5 years low-dose CT for 5 years
ITALUNG4 2004–06 Age 55–69 years, ≥20 PY smoker Annual low-dose CT for 4 years vs no screen
DANTE5 2001–06 Age 60–75 years, ≥20 PY smoker, quit smoking <10 years earlier, Annual low-dose CT for 4 years vs no screen
male
DLCST6 2004–06 Age 50–70 years, ≥20 PY smoker, quit smoking <10 years earlier, Annual low-dose CT vs usual care for 5 years
FEV1 ratio >30%, able to climb two flights of stairs without pausing
NELSON2 2003–06 Age 50–75 years, smoker or quit smoking ≤10 years earlier, Low-dose CT in year 1, year 2, year 4, and year 6·5 vs
>15 cigarretes per day for >25 years or >ten cigarretes per day for no screen
>30 years
LUSI7 2007–11 Age 50–69 years, heavy smoking history Annual low-dose CT and smoking cessation for
5 years vs smoking cessation alone
UKLS8 2011–14 Age 50–75 years, ≥5% of 5-year lung cancer risk as calculated by Wald single low-dose CT screen design vs no screen
LLPv2 scores
Other studies
I-ELCAP14 1993–2006 Age >60 years, ≥10 PY smoker Annual low-dose CT and chest x-ray for 5 years
Mayo LDCT trial18 1999 Age >50 years, 20 PY smoker, quit smoking <10 years earlier Annual low-dose CT for 5 years
PANCAN19 2008–11 Age 50–75 years, ≥2% of 3-year lung cancer risk as calculated by Low-dose CT in year 1, year 2, and year 4
PLCO score
COSMOS20 2000–01 Age >50 years, ≥20 PY smoker Annual low-dose CT for 10 years
PY=pack-year. FEV=forced respiration volume. LLPv2=Liverpool Lung Project risk model, version 2. PLCO=Prostate, Lung, Colorectal, and Ovarian trial risk model.
Table: European pilot trials for lung cancer low-dose CT screening

have lung cancer. This definition differs from that used
in the NLST, in which every individual who had an
additional CT scan before a repeat annual screen was
considered positive (a nodule with a diameter of 4 mm
or more). This amounted to 24% of participants, of
which 96% were false positive with unnecessary CT
examinations and a related radiotherapy burden.
Since the publication of NLST, the NELSON study21 has
shown that nodules with a volume smaller than
100 mm³ (or <5 mm in diameter) do not confer an
increased risk of malignancy at baseline.
Although no other technology is available that could
replace CT screening, emerging technologies need to
undergo the same scrutiny that has been applied to CT
screening to ensure robustness and validity in its use.
However, if a new emerging technology is considered, it
should be compared with CT screening in a randomised
controlled trial, and should show a negative predictive
value of almost 100% and a positive predictive value higher
than that of CT screening. In the future, some technologies
might be applied as an adjunct to CT screening, of which
more is discussed later on in this paper.
Outcomes of lung cancer screening trials
The outcomes of various lung cancer screening trials
provide insight into how to implement lung cancer
screening in differing countries in Europe, as well as the
optimal set-up for a population and screening at a single
centre. We have learnt much about each stage of the
lung cancer CT screening pathway and the management
decisions that are required.22 Ongoing trials2–8 have
provided insight into risk assessment, CT screen nodule
manage­ ment, multidisciplinary team work-up, and
surgical interventions, as well as psychological effects
on parti­ cipants and the cost-effectiveness of the
screening process.
Several nationally funded randomised studies have
already been undertaken in Europe to assess the
feasibility of lung cancer screening. These include
DANTE,5 DLCST,6 ITALUNG,4 LUSI,7 MILD,3,23 NELSON,2
and UKLS.8,22 Their results, individually and when
pooled, will contribute to the implementation of CT
screening in Europe. The only European fully powered
randomised controlled trial that will provide mortality
and cost-effectiveness data is the NELSON trial, but we
do have sufficient data to start planning before the full
results are available. The results from NLST alone have
been sufficient for low-dose CT screening to start in the
USA and Canada.
The incorporation of the coronary artery calcification
score and emphysema assessment on low-dose CT
imaging might enhance the cost-effectiveness and
attractiveness of low-dose CT lung cancer screening.24
Chronic obstructive pulmonary disease (COPD) and
emphysema are the strongest lung cancer risk predictors
and, together with cardiovascular disease, all three
imaging biomarkers have substantial effects on
www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6 3
Policy Review
morbidity but also have independent effects on overall
mortality.25,26
Lung cancer risk prediction modelling
The concept of clearly defining a target population for
lung cancer screening is gaining importance.19,27
Selection on the basis of age alone, as in most other
cancer screening disease settings (eg, breast and colon),
is insufficient in lung cancer because of other powerful
risk factors, the most important of which is exposure to
tobacco smoke. Other major risk factors, which are also
taken into account, include a history of respiratory
diseases (COPD, emphy­sema, bronchitis, pneumonia,
and tuberculosis), pre­vious malignancy, family history
of lung cancer (first-degree relative diagnosed at age
60 years or younger), and exposure to asbestos. Several
multivariable risk prediction models have been
published,28,29 but only two (the modified Liverpool Lung
Project [LLPv2] and Prostate, Lung, Colorectal and
Ovarian Cancer Screening Trial [PLCO] models27,30) have
been used so far to select patients for screening in a
clinical trial. Risk prediction models, including the
US Preventive Services Task Force (USPSTF) recommen­
dations, have been tested in the NLST dataset, showing
that the NLST selection criteria could have been
improved if a risk prediction model had been
implemented.30–32 The LLPv2 is the only risk model that
has been used so far to select patients for a lung cancer
screening randomised controlled trial.8 In this trial, a
higher percentage of participants were diagnosed with
lung cancer at baseline than those in the NLST and
NELSON trials. The LLPv2 model cutoff of 5% over
5 years is being validated in the Liverpool Healthy Lung
Project.27,33,34 The original LLP model has compared
favourably with the Spitz and Bach models.35 The LLP
model was validated in the UK with the use of data from
a population-based prospective cohort study,33 with an
area under the curve (AUC) of 0·82 (95% CI 0·80–0·85).33
The Bach, Spitz, LLP, and modified PLCO (PLCOm2012)
risk models were externally validated in the German
EPIC cohort study36 of 20  700 ever-smokers. The
PLCOm2012 model showed the best performance in
external validation (C index 0·81, 95% CI 0·76–0·86)
and the highest sensitivity, specificity, and positive
predictive value compared with the other three models.
However, the superiority of the PLCOm2012 model over the
Bach and LLP models was considered modest by the
authors.36
Five different risk models have been compared with one
another by use of the PLCO and NLST trial datasets.37
Although several sophisticated models have used a range of
risk variables (ie, family history, previous malignancy,
previous respiratory disease, exposure to asbestos), the
Bach model had a good sensitivity and specificity,28 even
though it only used age and smoking history for cal­culating
risk score, emphasising the importance of these two risk
factors. The PLCOm2012 model also provided good results,
 Policy Review
but this model was developed using the PLCO dataset, so
there might be issues of overfitting. However, all the
models compared were superior to the NLST selection
criteria and the USPSTF recommendations. The PLCOm2012
risk model was also used to calculate the predicted risk of
lung cancer from baseline data from 95 882 ever-smokers
aged 45 years or more in the Australian 45 and Up Study
(2006–09).38 The results showed good discrimination
(AUC 0·80, 95% CI 0·78–0·81) and excellent calibration.
Thus, the use of risk prediction models is essential in the
selection of patients for lung cancer screening. Cost-
effectiveness was also increased in the high-risk groups38
meaning that risk prediction should also reduce costs per
life saved. There is no information on related cost-
effectiveness.37 We recognise that the aforementioned risk
prediction models were based on European populations
and that lung cancer risk predictions might be affected by
regional differences. This EU position statement does not
recommend any specific risk prediction model, but either
the PLCOm2012 or the LLPv2 would suffice if screening were to
be implemented immediately.
A 80 mm3 100 mm3
5·35 mm 5·78 mm
5·35 mm
5·78 mm
B 268 mm3 524 mm3
8 mm 10 mm
8·0 mm
10·0 mm
Figure 1: Comparative visualisation demonstrating the advantage of using volume instead of diameter when
assessing CT-detected lung nodules
(A) A volume growth of 25%, defined as growth by NELSON criteria, is hardly appreciable by diameter
measurement (8% diameter increase, which is no growth according to existing criteria). (B) A 25% diameter
increase (ie, the threshold for growth definition) reflects almost a doubling in volume (95%), highlighting the
insensitivity of diameter measurement for growth. Reproduced with permission from Field and colleagues.22
4 www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6
We need to be aware of the different European health
care systems and the issues concerning the use of a risk
stratification approach (such as in Germany, in which all
individuals have a legal right of access to the available
diagnostic and therapeutic techniques). However, it
should be argued that screening low-risk patients would
be unethical because of harm-to-benefit considerations.
The risk profile of those targeted for screening is a
valuable and cost-effective tool to identify those with
preclinical disease who are eligible for screening.8,30 The
integration of the risk profile with one or more
biomarkers or susceptibility genes could improve the
selection of high-risk patients for screening, the manage­
ment of disease, or both.39,40 Predictive biomarkers, such
as microRNA markers, are potentially effec­tive tools for
the identification of susceptible patients and future lung
cancer cases,41–43 while a bronchial airway gene-expression
classifier could improve the diagnostic per­formance of
bronchoscopy.42 Breath tests for lung cancer should be
considered a strong possibility for screening and are
being tested in clinical trials.43,44
The identification of new biomarkers for screening will
be a reason to implement cooperative research. The
availability of large, high-quality biobanks embedded in
screening trials together with radiomic analysis is a
future opportunity that should be explored further in a
lung cancer screening context.
Harms and benefits associated with lung cancer
screening
The harms associated with lung cancer screening, such
as overdiagnosis, surgery for benign lesions,
psychological harm, and radiation exposure, need to be
acceptable before the implementation of screening.
Minimising harm is essential to maximise the clinical
effectiveness of the intervention.
Physical harms can be reduced by ensuring that only
patients with a sufficiently high risk of developing lung
cancer are screened, by reducing screening radiation
dose to a minimum, and by the effective management of
atypical findings, including nodules, suspected lung
cancers, and incidental findings. These measures re­
quire a high degree of clinical expertise to be available so
that all aspects of CT screening and management are
completed to the highest standards. Thus, lung cancer
screening should only be undertaken according to
protocol, and at screening units and centres that are able
to guarantee rigorous quality control.
According to the guidelines of the European Society for
Medical Oncology (ESMO)45 and the European Society of
Thoracic Surgeons,46 low-dose CT screening can be done
outside a clinical trial setting if it is offered within a
screening programme with quality control and  in a
centre with CT screening experience, extensive thoracic
oncology activity, multidisciplinary management of
suspicious findings, and a well-developed programme of
minimally invasive thoracic surgery. Psychological harms
 Policy Review

can be reduced by providing information about CT An alternative method is to determine nodule volume
screening in a language that is understood by those who using a software for semi-automated segmentation, which
are screened, including details about atypical findings, enables an accurate estimation of nodule size after
with accurate information about the probability of cancer,
especially where findings are likely to be benign.
The potential physical harms of screening should be 14
provided to attendees in a clear manner, including harm
from radiation exposure47 and the harms from a biopsy or *
12
resection of a benign lesion. However, the radiation risk * *
is likely to be overestimated and will decrease in
the future with the arrival of the latest CT platforms, with 10 *
ultra low-dose CT technology. All European trials will *
**
Mean axial diameter (mm)
provide data that will allow a direct quantification of
overdiagnosis. The proportion of benign resections in 8
clinical trials varies from 10% to at least 25% of all
operations.3,8 Our consensus indicates that a prevalence
6
of 10% or lower for lung cancer should be reached, but
an optimal percentage has not been established yet. It
should be noted that the dynamic between patient and 4
physician is altered in a lung cancer screening setting
when compared with settings where symptomatic
2
individuals present them­selves to health-care institutions.
Effective implementation of lung cancer screening
programmes also includes recognition of the benefits of 0
maximising smoking cessation within CT screening 0 50–100 100–200 200–300 300–400 400–500
Nodule volume categories (mm3)
programmes. Smokers should be informed of the dangers
of continuing to smoke for their own general health and Figure 2: Range in mean axial nodule diameter per nodule category
should be offered suitable support to help quit.48–50 Nodules with a mean diameter of 8–10 mm (coloured zone) are represented in each nodule volume category.
These nodules represent the group with the highest uncertainty about nodule nature. The data in this figure are
based on intermediate-sized baseline nodules only. Adapted with permission from Heuvelmans and colleagues­.58
CT methodologies for early lung cancer
detection
In the NLST trial, a CT screen was regarded as positive if
Solid non-calcified nodule at baseline CT
it showed any non-calcified nodule of at least 4 mm in
diameter. The American College of Radiology set up a
Yes
Lung Cancer Screening Committee subgroup to develop Clear features of benign disease?
Lung-RADS,51,52 a quality assurance tool with which to
No
standardise the reporting of lung cancer CT screening
and to inform management recommendations. The Volumetric analysis (or diameter measurement if
rationale behind this initiative was that it would assist in volumetry not available or not technically possible)

the interpretation of nodule findings.

A comparison of Lung-RADS performance with NLST
data53 showed that Lung-RADS substantially reduced the <100 mm³ volume or 100 to <300 mm³ volume ≥300 mm³ volume or
false-positive result rate, but also reduced screening <5 mm diameter or 5 to <10 mm diameter ≥10 mm diameter
sensitivity. Mehta and colleagues54 have suggested that the
Lung-RADS system needs to be revised, and they faulted
CT scan 3 months after
the system on the basis that it has never been studied in a baseline
prospective manner. Additionally, Li and colleagues55 have
analysed the effect of the so-called rounding method used
in Lung-RADS on the frequency of positive results and on Next round of screening No VDT ≤600 days? Yes Further work-up and
according to protocol consideration of
the growth assessment of pulmonary nodules. The authors definitive management
con­cluded that rounding up the mean nodule diameter in
Lung-RADS increased the frequency of positive results,
Management according
leading to a detrimental effect on the efficiency of lung to category at 3 months
cancer screening. Furthermore, Lung-RADS does not
provide guidance on risk prediction models. The Brock Figure 3: Nodule management protocol for screen-detected solid nodules at baseline
score provides a more accurate estimate of a nodule’s risk For nodules with a volume-doubling time (VDT) of 400–600 days (intermediate cancer risk of about 4%), a second
of malignancy than baseline Lung-RADS criteria.56,57 repeat CT scan in 3 months should be considered as an initial work-up option.

www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6 5

 Policy Review

three-dimensional reconstruction (figure 1). Volumetric
analysis of CT-detected nodules was initially recommended
by Henschke and colleagues14 in 1999, and has been
further developed and validated in the NELSON and the
UKLS trials. A comparative analysis58 of both diameter
and volume was done with baseline data from participants
of the NELSON trial, of which 2240 non-calcified nodules
of intermediate size were identified. Diameter within a
single nodule varied by a median of 2·8 mm, which is
larger than the LungRADS cutoff for nodule growth
(>1·5 mm increase in mean diameter). Nodules with a
diameter of 8–10 mm were represented in each of the
five nodule volume categories (figure 2).59
The recommendation for the future management of
solid nodules detected with CT screening is that semi-
automatically derived volume and volume-doubling time
should be used in preference to diameter measurements,
which should only be used where volumetry is not
technically possible.
Prerequisites for lung cancer population
screening
The accreditation awarded to institutions and radiologists
participating in lung cancer CT screenings should
include training in the implementation of quality
assurance processes.
The establishment of central national registries for
participants would ensure that inclusion criteria are met.
In this registry, results from different screening
modalities, such as CT manufacturer dose, together with
work-up results, should be collected to ensure that
previous screens are available and quality control can be
assured. The institutions providing a lung cancer
screening service should be registered, have access to a
participant registry that includes information from
previous screens, should use a certified nodule evaluation
software, and should deliver screening results and
recommendations to a central participant registry. We
recommend that the European lung cancer community
develop national registries, which could be linked on a
hub-and-spoke model, to enable international quality
con­trol and the use of collected data to improve the
provision of lung cancer screening throughout Europe
over time.
National quality assurance boards should be set up to
monitor the adherence to minimum technical standards
and to standardise diagnostic criteria for screen-detected
lung nodules, similar to the UK and European breast

Newly identified solid non-calcified nodule

not present on the previous CT screening

Yes Next round of screening

Clear features of benign disease?
according to protocol
No

Volumetry (or diameter measurement if volumetry

is not available or not technically possible)

<30 mm³ volume or 30 to <200 mm³ volume ≥200 mm³ volume

<4 mm diameter or 4 to <8 mm diameter or ≥8 mm diameter

CT scan 3 months after Further work-up and

detection consideration of
definitive management

Nodule resolution, benign Stable size on basis of VDT >600 days and VDT ≤600 days or
calcification, or significantly volumetry or two- <200 mm³ volume or ≥200 mm³ volume or
decreased size dimensional non-automated <8 mm diameter ≥8 mm diameter
diameter value

Next round of screening according to protocol

Management according to category at 3 months

Figure 4: Nodule management protocol for screen-detected incidental solid nodules at follow-up
VDT=volume doubling time.

6 www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6

 Policy Review

Baseline volumetric analysis (or diameter measurement if

volumentric is not available or not technically possible)

5–6 mm diameter ≥80 mm³ volume or

≥6 mm diameter

CT scan 3 months
after baseline

No Yes
CT scan 1 year after VDT ≥400 days or clear
baseline evidence of growth

Stable on basis of two- Stable size on basis of VDT >600 days VDT 400–600 days VDT ≤400 days or clear
dimensional non- volumetry evidence of growth
automated diameter value

CT scan 2 years after Discharge Consider discharge (only
baseline if based on volumetry)
or ongoing CT
surveillance depending
on patient preference
Consider biopsy or Further work-up and
further CT surveillance consideration of
depending on patient definitive management
preference

VDT assessment and

manage according to
VDT category at 1 year;
discharge if stable

Figure 5: Nodule management protocol for clinically detected solid nodules according to British Thoracic Society guidelines
VDT=volume doubling time. Reproduced with permission from Callister and colleagues.64

screening programmes.60–62 Such national quality assurance
boards should be entitled to advise or intervene whenever
basic requirements are not met. The lung cancer
community should consider following the example of the
Dutch breast cancer screening service by organising so-
called central reading centres of all CT screening
programmes across the country.62 This ap­ proach is
favoured over a local reading of CT scans as the latter could
have a major effect on routine radiology service delivery.
This approach would also enable ongoing national quality
assurance control and the introduction of the forefront
automated pulmonary nodule reading software.
Institutions participating in screening programmes
require multidisciplinary teams to represent all relevant
specialities (including a pulmonologist, thoracic surgeon,
radiologist, lung cancer nurse, and so on) in which
suspicious screening results can be discussed. These
institutions should regularly demonstrate to a quality as­
surance board that they continue to meet basic standards
similar to those proposed by the Radiological Society of
North America.63
Lung nodule management at baseline CT
screening
The management of prevalent lung nodules should
mostly depend on size criteria. Volumetry is essential, but
diameter cutoffs need to be provided for cases in which
segmentation is not possible. Minimum standards for CT
acquisition parameters in lung cancer screening need to
be met to ensure the standardisation of volumetric
analysis (ie, protocol regarding slice thickness, recon­
struction interval, and image reconstruction algorithm
[kernel]), and to clearly define the low radiation dose.
The management of screen-detected lung nodules
should be based on the evidence from screening trials
that have used volumetry, such as the NELSON trial. In
the original NELSON nodule management protocol,2
volume cutoffs for negative and positive screen results
were less than 50 mm³ for negative and more than
500 mm³ for positive results. Nodules with a volume of
50–500 mm³ were classified as indeterminate. These
cutoffs could be optimised on the basis of lung cancer
probability results of the first two screening rounds from
the NELSON trial.21 For example, for solid nodules with a
volume of less than 100 mm³, the patient should return
for an annual screen; for nodules with a volume of
100–300 mm³, the patient should return for a repeat
screen in 3 months; for volumes greater than 300 mm³,
the patient should be referred to a multi­ disciplinary
team.21 Figure 364 shows the recom­ mended nodule
management protocol for screen-detected solid nodules
at baseline, figure 464 for screen-detected incidental

www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6 7

 Policy Review

nodules at follow-up, figure 5 for clinically detected solid
nodules according to British Thoracic Society (BTS)
guidelines, and figure 6 for sub-solid nodules for both
screen-detected and clinically detected nodules. Detailed
risk profiles for the probability of lung cancer over 2 years
are shown in figure 7.21 Data to inform this figure have
been provided by the NELSON group for both nodule
volume and volume-doubling time (<400 days and
400–600 days—an increased risk, described in figure 3;
>600 days, no substantially increased risk), which
provides guidance to the future follow-up interval for
specific participants. In 2017, a study65 provided in-vivo
evidence for the growth patterns of screen-detected lung
cancers, showing an exponential growth pattern that can
be described by the volume-doubling time. Acknowledging
that software packages give different estimates of solid
nodule volume, commonly around 20% of difference
(corresponding to a non-measurable 7% error in nodule
diameter; absolute 0·4 mm error),66 there might be merit
in decreasing the nodule volume threshold for a repeat
screen at 3 months to 80 mm³ if the software is not
phantom validated (a calibration process for quality
assurance of different CT scanners; figure 5).
For sub-solid nodules, surveillance should be favoured
over intervention to avoid overdiagnosis. For all pure
ground glass nodules and most partial solid nodules,
a return to annual screening is recommended (figure 6).67
Knowledge and data from ongoing lung cancer screening
projects will also be important for the future optimisation
and refinement of nodule management protocols.
Morphology assessment should also play a part in
the management of solid nodules, such as clustered,
ill-defined nodules, which are more consistent with
inflammatory aetiologies, or smooth peri-fissural
nodules or intrapulmonary lymph nodes, which require
management not based purely on size criteria.68 There
are several alternative work-up methods for screen-
detected suspicious nodules with volumes larger than
300 mm³ at baseline, such as core needle biopsy, PET
or CT scans, and primary resection.
The management of a patient should be done ac­
cording to their risk of malignancy. Low-risk nodules,
such as those with a risk of malignancy lower than 10%,
can be followed up with interval imaging, but high-risk
nodules need further work-up if it is agreeable to the
patient after an informed discussion. As the risk of

Sub-solid nodule on CT

Yes
Nodule <5 mm, patient unfit for any
treatment or stable over 4 years?

No
Yes
Previous imaging? Assess interval change; if stable over
less than 4 years, assess risk of
No malignancy as below
Repeat thin section CT at 3 months

Resolved Stable Growth or altered morphology*

Assess risk of malignancy

(Brock model†, morphology‡),
patient fitness, and patient
preference

Low risk of malignancy Higher risk of malignancy

(approximately <10%) (approximately >10%) or concerning
morphology‡, discuss
options with patient

Discharge Thin section CT 1, 2, and 4 years Image-guided biopsy Favour resection or non-surgical
from baseline treatment

Figure 6: Management protocol for sub-solid nodules for both screen-detected and clinically detected nodules according to British Thoracic Society
guidelines
Reproduced with permission from Callister and colleagues.64 *Change in mass or a new solid component. †The Brock model can underestimate the risk of malignancy
in sub-solid nodules that persist at 3 months. ‡The size of the solid component in part-solid nodules, pleural indentation, and bubble-like appearance.

8 www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6


malignancy increases, management options broadly
include further surveillance, biopsy, or treatment.
ESMO guidelines published in 201745 indicate that the
cornerstone of treatment of potentially resectable lung
cancer is the surgical removal of the tumour. For patients
who are not willing to accept the risks of surgery, or for
whom surgery is a high-risk option, non-surgical therapy
should be offered. This option could be either stereo­
tactic ablative radiotherapy, hypofraction­ated high‑dose
radiotherapy or image-guided ablative therapy.45
Incident screening rounds
Although incident screening rounds will constitute much
of the work in the early detection of lung cancer, until
recently research did not focus on incident nodules and
their definition, which has varied widely between low-
dose CT lung cancer screening trials.8,64,69,70 Incident
nodules detected in high-risk individuals after baseline
screening had either been missed in a previous scan or
had developed de novo in the time interval since the
previous scan. In the event of a missed nodule, calculation
of the volume-doubling time is advised for further risk
stratification. However, patients with newly developed
nodules are a specific group that is distinct from patients
that have had nodules detected at baseline. With an
annual incidence of between 3% and 13% of participants,
these newly developed nodules are regularly encountered
in low-dose CT lung cancer screening.71–74 Unlike baseline
nodules, which could have been present for years before
detection, new incident nodules are potentially fast-
growing.17,75–77 This potential is reflected in a high cancer
risk of 2–8% for participants with a new incident
nodule.17,71,72,74 Because these nodules have less time to
grow before detection than those nodules detected
previously, baseline cutoff values are not applicable.17 This
formerly theoretical concept, which led to an adjustment
of cutoff values for new incident nodules in several
trials,53,72,77 has been supported for new solid incident
nodules by the results of the NELSON trial.17 Because a
large proportion (37–57%) of new incident nodules are
very small (<50 mm³ in volume),17,71,74 volume measure­
ments should be preferred, since diameter measurements
are much less precise and reproducible. Data from the
NELSON trial suggest that new solid incident nodules
should be categorised in three groups: nodules smaller
than 27 mm³ in volume (<1% lung cancer probability)
represent a low-risk group, and these patients could
return to the annual screen schedule (based on an annual
screening programme); patients with new solid incident
nodules of 27–207 mm³ in volume (3% lung cancer
probability) form an intermediate-risk group requiring a
repeat low-dose CT in 3 months; and patients with new
non-calcified solid incident nodules equal or greater than
208 mm³ in volume (17% lung cancer probability) form a
high-risk group requiring referral to a multidisciplinary
team.17 We  suggest simplifying these categories to
volumes smaller than 30 mm³, 30–200 mm³, and equal or
www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6 9
Policy Review
700
600
500
Volume doubling time (days)
400
3%
300
200
5%
10% 15% 20% 30% 50%
7%
100
0
0 100 200 300 400 500
Volume of largest nodule (mm3)
Figure 7: Contour plot of the effect of the combined effect of nodule volume and volume-doubling times on
the 2-year probability of lung cancer
The risk isolines represent the percentage of NELSON participants that will be diagnosed with lung cancer within
2 years according to the volume of their largest nodule and volume-doubling time of the fastest growing nodule in
the 50–500 mm3 range. Reproduced with permission from Horeweg and colleagues.21
greater than 200 mm³ (figure 4). The existing data indicate
that the majority (68–86%) of lung cancers found in new
incident nodules during lung cancer screening are
detected at stage I.17,72 Therefore, volume-doubling time
assessment at follow-up scans appears appropriate, as
outlined in the BTS guidelines.64 However, the available
evidence regarding new incident nodules is insufficient,
and a more standardised approach to reporting, such as
strictly separating baseline and incident nodules, could
simplify the recommended routine clinical management
of patients with newly detected incident nodules. If a CT
scan is done less than 2 years before screening is available,
recommendations for new inci­ dence nodules detected
during screening could be extrapolated to routine clinical
practice in a high-risk patient population, similar to that
done in the NELSON trial. This recommendation has
now been adopted by the BTS guidelines on nodule
management,78 and by the BTS Quality Standard on lung
nodule management. In a low-risk population, the
management of patients should follow BTS guidelines.
Clinical work-up of CT-detected lung nodules in
clinical practice
Incidentally detected lung nodules are an increasingly
common clinical problem arising from the increased use
of cross-sectional imaging in clinical practice. The BTS
has undertaken the in-depth task of developing guide­lines
on the management of pulmonary nodules in a clinical
context, separate from the context of population
screening.64 This work has been based on an extensive
 Policy Review
Panel 1: European Union (EU) position statement recommendations
We recommend the following actions to begin implementation of lung cancer screening
in Europe:
1 Low-dose CT is the only evidence-based method for the early detection of lung cancer
shown to provide a mortality reduction. On the basis of this evidence from
randomised controled trials, the EU position statement recommends that we start to
plan for the implementation of lung cancer screening in Europe while cognisant of
future publications that include the awaited NELSON trial data on mortality and
cost-effectiveness and data from the six smaller European studies for developing
implementation strategies in each of their own countries.
2 Future lung cancer low-dose CT programmes should use a validated risk stratification
approach so that only individuals deemed to be at high enough risk are screened. In
the near future, incorporation of potential biomarkers and susceptibility genes into
lung cancer risk models should be considered to improve the accuracy of risk
stratification models.
3 All future screenees entering into early detection programmes for lung cancer should
be provided with carefully constructed participant information on the potential
benefits and harms of screening to enable them to make an informed decision as to
whether they wish to participate or not. Smoking cessation advice should be offered
to all active smokers.
4 Future management of screen-detected solid nodules should utilise semi-automatically
derived volume measurements and volume-doubling time, and should be quality
assured.
5 National quality assurance boards should be set up by professional bodies to ensure
adherence to all minimum technical standards, including semi-automated volumetry,
and to standardise diagnostic criteria for screen-detected lung nodules, including
radiation exposure limits.
6 Management of prevalent lung nodules in CT screening programmes, lung nodules at
incident screening (newly detected), and CT-detected lung nodules in clinical practice
should be managed with different protocols because of different pretest lung cancer
probabilities.
7 Although only evidence for annual low-dose CT lung cancer screening is available,
recent research suggests the possibility of using a more personalised approach to lung
cancer screening with a risk-based approach on the results of baseline and first
screening rounds.
8 Management of lung nodules by lung cancer multidisciplinary teams should be done
according to the EU position statement recommendations with the aim of minimising
harm and ensuring patients receive the most appropriate treatment.
9 The EU position statement expert group recommends that the planning for low-dose
CT screening should be started throughout Europe because low-dose CT lung cancer
screening has the potential to save lives.
review of the literature, which included publications from
several lung cancer CT screening trials, and an in-depth
analysis of the data. The guideline development group
used methods compliant with the AGREE Collaboration
criteria and standards set by NHS Evidence. The evidence
review was comprehensive, done in November, 2012, and
updated in June, 2014. The guidelines provide four
management algorithms and two malignancy prediction
tools:64 the Brock risk prediction tool to calculate
malignancy in solid pul­monary nodules equal or larger
than 5 mm in volume, which are unchanged at 3 months,56
and the Herder prediction tool to be used after PET/CT.79
Furthermore, volumetry has been recommended by BTS
10 www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6
as the preferred measurement for CT-detected nodules.
The guidelines also provide recommendations for the
management of nodules with extended volume-doubling
times.
The BTS guidelines provide recommendations on the
use of further imaging and PET/CT information that can
be incorporated into pulmonary risk models (Herder
model), and advice on biopsy and the threshold for
treatment without histological confirmation. BTS also
provides advice on the information that should be given
to patients about the management of pulmonary nodules
in a non-screening context. This EU position statement
recommends keeping a database of all nodules that can
facilitate a future refinement of nodule management in
line with new evidence.
Optimal timing of lung cancer screening
intervals
The USPSTF on CT screening has recommended annual
screening from age 55 years to 80 years.80 In a NELSON
publication,81 a 2·5-year screening interval resulted in a
significant increase in in­ terval cancers in the fourth
screening round, providing evidence against using this
interval in a future screening programme. There were also
significantly more interval cancers in a 2·5-year timeframe,
and a trend towards more cancers detected at a later stage.
A detailed analysis of the cost-effectiveness of various
screening scenarios showed that almost all approaches
increase cost-effectiveness when screens are annual.82
However, half of the participants included in the NELSON
trial had no pulmonary nodules detected, and the 2-year
probability of participants developing lung cancer was
0∙4%, indicating that a screening interval of up to 2 years
could be considered for similar individuals in future
screening programmes, in a risk-stratified approach. The
only trial to test annual and biennial screening was the
MILD trial,83 in which no difference was found in terms of
mortality when comparing these two screening intervals.
Screening intervals have been modelled by both the
UKLS trial84 and the International Early Lung Cancer
Action Program.85 Duffy and colleagues84 acknowledged
the risk of increasing the number of screening intervals
but also acknowledged that it could potentially provide a
more cost-effective approach. Yankelevitz and colleagues85
argued that we should move beyond hypothesis testing
and onto quantification. We need to learn how the length
of the interval between screens affects the diagnostic
distribution before we consider changing annual screening
intervals.
So far, we only have trial evidence for annual screening.
Studies have shown that previous negative screening
results might provide directions for further risk
stratification.86,87 Future decisions regarding interval
timing should be based on risk, psychosocial effects,88
cost-effectiveness, and the feasibility of implementation,89
but these areas require further investigation. However,
with new ultra-low-dose CT techniques, the radiation

Panel 2: A call for action
Europe needs to set a timeline for implementing lung cancer
screening:
• Publish recommendations for implementation with
quality assurance measures (6 months)
• Plan health service requirements and their delivery
(12 months)
• Plan phased implementation in high-risk regions while
awaiting mortality data from the NELSON trial (18 months)
• Plan to set up a European registry of images and data
(18 months)
• Evaluate implementation after the first 12 months and
review delivery strategy (36 months)
• Expand lung cancer screening to all regions within Europe
(48 months)
dose for repeated CT screenings over a 30-year period
might not be a major issue for participants. New
developments, such as deep machine learning, will assist
in the automation of pulmonary nodule management in
lung cancer screening.90
In the future, we think that, with implementation of
ultra-low-dose CT screening, there will be no obstacles in
tailoring the frequency of screening of high-risk
individuals over a 25-year period. We should be
considering precision medicine in the field of lung
cancer screening, and whether an individual who has
had a negative baseline and a negative 1-year scan should
be moved into biennial screening until their risk profile
changes. As lung cancer screening is still in an embryonic
stage of implementation in Europe, we have an op­
portunity to plan the development of an optimal lung
cancer low-dose CT screening strategy.91
Conclusion
This EU position statement describes the current status of
lung cancer screening in Europe. Through consensus
discussions with experts from the eight European
countries undertaking randomised controlled trials of
lung cancer CT screening, we have developed nine
recommendations to guide the implementation of lung
cancer screening in Europe (panel 1). Some specific areas
still require further development and consideration, such
as the integratation of smoking cessation into lung cancer
screening programmes and the selection of the appropriate
target screening population. However, evidence clearly
shows that Europe must start planning for implementation
within the next 18 months, as outlined here (panel 2).
During this planning period, each country will need to
focus on deciding the best risk prediction method for the
identification and recruitment of high-risk populations
and on setting up the required infrastructure for
quality-controlled CT scans that use volumetric analysis.
This EU position statement has provided detailed
recommendations on the management of lung nodules by
www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6 11
Policy Review
Search strategy and selection criteria
Data for this European Union position statement were
identified through searches of PubMed, MEDLINE, and
references from relevant articles using search terms “lung
cancer CT screening trial”, “lung screen detected nodules”,
“lung cancer CT screening recommendations”, and “lung
cancer CT screening cost effectiveness”. Identified abstracts
and reports from meetings were included only when they
related directly to previously published work. Only articles
published in English between 1999 and 2017 were included.
lung cancer multidisciplinary teams, with the aim of
minimising harm and ensuring that patients receive the
optimal diagnosis and therapy.
Contributors
MO and JKF developed the concept and design of the EU position
statement on lung cancer screening. All authors contributed equally
to the development of the EU position statement.
Declaration of interests
GB has received personal fees from Bayer, General Electric, and Siemens
Healthcare. DRB has received personal fees from AstraZeneca. SD has
received grants from the German Research Foundation and from the
Dietmar Hopp Foundation. JKF has received grants from HTA funding
for the UKLS trial, grants and other funding from Liverpool CCG, and
other research funding from Epigenomics and Vision Gate. CPH has
received consultation and personal fees from Pfizer, Boehringer
Ingelheim, Novartis, Gilead, MSD, Intermune, and Fresenius; research
funding from Siemens, Pfizer, and Boehringer Ingelheim; and lecture
fees from Gilead, MSD, Pfizer, Intermune, Novartis, Basilea, and Bayer.
MVI reports personal fees from Exact Sciences. HdK reports grants and
other non-financial support from the NELSON trial, a grant for health
technology assessment for CT lung cancer screening in Canada
by Cancer Care Ontario, and a grant from the University of Zurich
to assess the cost-effectiveness of CT lung cancer screening in
Switzerland; HdK took part in a 1-day advisory meeting on biomarkers
organised by MD Anderson/Health Sciences during the 16th World
Conference on Lung Cancer. MO reports grants from the Royal Dutch
Academy of Sciences, from the Netherlands Organisation of Scientific
Research, and from the Netherlands Organisation for Health Research
and Development. NS reports personal fees from Roche, Boehringer
Ingelheim, Parexel, and Bayer. WR reports a patent of a protein marker
signature of early lung cancer pending, and a patent of a miRNA
signature of early lung cancer. AD, RV, TH, HP, MM, MEC, MAH, NB,
UP, JHP, EP, and SWD declare no competing interests.
References
1 National Lung Screening Trial Research Team, Aberle DR,
Adams AM, et al. Reduced lung-cancer mortality with low-dose
computed tomographic screening. N Engl J Med 2011; 365: 395–409.
2 van Klaveren RJ, Oudkerk M, Prokop M, et al. Management of lung
nodules detected by volume CT scanning. N Engl J Med 2009;
361: 2221–29.
3 Sverzellati N, Silva M, Calareso G, et al. Low-dose computed
tomography for lung cancer screening: comparison of performance
between annual and biennial screen. Eur Radiol 2016; 26: 3821–29.
4 Paci E, Puliti D, Lopes Pegna A, et al. Mortality, survival and
incidence rates in the ITALUNG randomised lung cancer screening
trial. Thorax 2017; 72: 825–31.
5 Infante M, Cavuto S, Lutman FR, et al. Long-term follow-up results
of the DANTE trial, a randomized study of lung cancer screening
with spiral computed tomography. Am J Respir Crit Care Med 2015;
191: 1166–75.
6 Wille MM, Dirksen A, Ashraf H, et al. Results of the randomized
danish lung cancer screening trial with focus on high-risk profiling.
Am J Respir Crit Care Med 2016; 193: 542–51.
 Policy Review
7 Becker N, Motsch E, Gross ML, et al. Randomized study on early
detection of lung cancer with MSCT in Germany: results of the first
3 years of follow-up after randomization. J Thorac Oncol 2015;
10: 890–96.
8 Field JK, Duffy SW, Baldwin DR, et al. The UK Lung Cancer
Screening Trial: a pilot randomised controlled trial of low-dose
computed tomography screening for the early detection of lung
cancer. Health Technol Assess 2016; 20: 1–146.
9 Field JK, van Klaveren R, Pedersen JH, et al. European randomized
lung cancer screening trials: Post NLST. J Surg Oncol 2013;
108: 280–86.
10 Van den Bergh KA, Essink-Bot ML, Bunge EM, et al. Impact of
computed tomography screening for lung cancer on participants in
a randomized controlled trial (NELSON trial). Cancer 2008;
113: 396–404.
11 Van den Bergh KA, Essink-Bot ML, Borsboom GJ, et al. Short-term
health-related quality of life consequences in a lung cancer CT
screening trial (NELSON). Br J Cancer 2010; 102: 27–34.
12 Fontana RS, Sanderson DR, Woolner LB, Taylor WF, Miller WE,
Muhm JR. Lung cancer screening: the Mayo program. J Occup Med
1986; 28: 746–50.
13 Oken MM, Hocking WG, Kvale PA, et al. Screening by chest
radiograph and lung cancer mortality: the prostate, lung,
colorectal, and ovarian (PLCO) randomized trial. JAMA 2011;
306: 1865–73.
14 Henschke CI, McCauley DI, Yankelevitz DF, et al. Early lung cancer
action project: overall design and findings from baseline screening.
Lancet 1999; 354: 99–105.
15 Biederer J, Ohno Y, Hatabu H, et al. Screening for lung cancer: does
MRI have a role? Eur J Radiol 2017; 86: 353–60.
16 Horeweg N, van der Aalst CM, Vliegenthart R, et al.
Volumetric computed tomography screening for lung cancer:
three rounds of the NELSON trial. Eur Respir J 2013; 42: 1659–67.
17 Walter JE, Heuvelmans MA, de Jong PA, et al. Occurrence and lung
cancer probability of new solid nodules at incidence screening with
low-dose CT: analysis of data from the randomised, controlled
NELSON trial. Lancet Oncol 2016; 17: 907–16.
18 Marcus PM, Bergstralh EJ, Fagerstrom RM, et al. Lung cancer
mortality in the Mayo Lung Project: impact of extended follow-up.
J Natl Cancer Inst 2000; 92: 1308–16.
19 Tammemagi MC, Schmidt H, Martel S, et al. Participant selection
for lung cancer screening by risk modelling (the Pan-Canadian
Early Detection of Lung Cancer [PanCan] study): a single-arm,
prospective study. Lancet Oncol 2017; 18: 1523–31.
20 Veronesi G, Bellomi M, Mulshine JL, et al. Lung cancer
screening with low-dose computed tomography: a non-invasive
diagnostic protocol for baseline lung nodules. Lung Cancer 2008;
61: 340–49.
21 Horeweg N, van Rosmalen J, Heuvelmans MA, et al. Lung cancer
probability in patients with CT-detected pulmonary nodules:
a prespecified analysis of data from the NELSON trial of low-dose
CT screening. Lancet Oncol 2014; 15: 1332–41.
22 Field JK, Oudkerk M, Pedersen JH, Duffy SW. Prospects for
population screening and diagnosis of lung cancer. Lancet 2013;
382: 732–41.
23 Infante M, Sestini S, Galeone C, et al. Lung cancer screening with
low-dose spiral computed tomography: evidence from a pooled
analysis of two Italian randomized trials. Eur J Cancer Prev 2016;
26: 324–29.
24 Cressman S, Peacock SJ, Tammemagi MC, et al.
The cost-effectiveness of high-risk lung cancer screening and
drivers of program efficiency. J Thorac Oncol 2017; 12: 1210–22.
25 Gonzalez J, Marin M, Sanchez-Salcedo P, Zulueta JJ. Lung cancer
screening in patients with chronic obstructive pulmonary disease.
Ann Transl Med 2016; 4: 160.
26 Murray CJ, Vos T, Lozano R, et al. Disability-adjusted life years
(DALYs) for 291 diseases and injuries in 21 regions, 1990–2010:
a systematic analysis for the Global Burden of Disease Study 2010.
Lancet 2012; 380: 2197–223.
27 Cassidy A, Myles JP, van Tongeren M, et al. The LLP risk model:
an individual risk prediction model for lung cancer. Br J Cancer
2008; 98: 270–76.
28 Bach PB, Kattan MW, Thornquist MD, et al. Variations in lung
cancer risk among smokers. J Natl Cancer Inst 2003; 95: 470–78.
12 www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6
29 Spitz MR, Hong WK, Amos CI, et al. A risk model for prediction of
lung cancer. J Natl Cancer Inst 2007; 99: 715–26.
30 Tammemagi MC, Katki HA, Hocking WG, et al. Selection criteria
for lung-cancer screening. N Engl J Med 2013; 368: 728–36.
31 Kovalchik SA, Tammemagi M, Berg CD, et al. Targeting of low-dose
CT screening according to the risk of lung-cancer death.
N Engl J Med 2013; 369: 245–54.
32 Katki HA, Kovalchik SA, Berg CD, Cheung LC, Chaturvedi AK.
Development and validation of risk models to select ever-smokers
for CT lung cancer screening. JAMA 2016; 315: 2300–11.
33 Raji OY, Duffy SW, Agbaje OF, et al. Predictive accuracy of the
Liverpool Lung Project risk model for stratifying patients for
computed tomography screening for lung cancer: a case-control and
cohort validation study. Ann Intern Med 2012; 157: 242–50.
34 Field JK, Gaynor ES, Duffy SW, et al. Liverpool healthy lung project:
a primary care initiative to identify hard to reach individuals with a
high risk of developing lung cancer. AACR Annual Meeting;
Washington, DC, USA; April 1–5, 2017. 4220.
35 Field JK, Marcus M, Maroni R, et al. Liverpool Healthy Lung
project. 2017. http://www.liverpoolccg.nhs.uk/health-and-services/
healthy-lungs/ (accessed June 9, 2017).
36 Li K, Husing A, Sookthai D, et al. Selecting high-risk individuals for
lung cancer screening: a prospective evaluation of existing risk
models and eligibility criteria in the german EPIC cohort.
Cancer Prev Res (Phila) 2015; 8: 777–85.
37 Ten Haaf K, Jeon J, Tammemagi MC, et al. Risk prediction models
for selection of lung cancer screening candidates: A retrospective
validation study. PLoS Med 2017; 14: e1002277.
38 Weber M, Yap S, Goldsbury D, et al. Identifying high risk individuals
for targeted lung cancer screening: Independent validation of the
PLCOm2012 risk prediction tool. Int J Cancer 2017; 141: 242–53.
39 Hung RJ, McKay JD, Gaborieau V, et al. A susceptibility locus for
lung cancer maps to nicotinic acetylcholine receptor subunit genes
on 15q25. Nature 2008; 452: 633–37.
40 Amos CI, Dennis J, Wang Z, et al. The OncoArray consortium:
a network for understanding the genetic architecture of common
cancers. Cancer Epidemiol Biomarkers Prev 2017; 26: 126–35.
41 Sestini S, Boeri M, Marchiano A, et al. Circulating microRNA
signature as liquid-biopsy to monitor lung cancer in low-dose
computed tomography screening. Oncotarget 2015; 6: 32868–77.
42 Silvestri GA, Vachani A, Whitney D, et al. A bronchial genomic
classifier for the diagnostic evaluation of lung cancer. N Engl J Med
2015; 373: 243–51.
43 Broza YY, Kremer R, Tisch U, et al. A nanomaterial-based breath
test for short-term follow-up after lung tumor resection.
Nanomedicine 2013; 9: 15–21.
44 Owlstone Medical. LuCID: a multi-centre prospective trial for lung
cancer screening. 2016. https://www.owlstonemedical.com/clinical-
pipeline/lucid/ (accessed July 1, 2017).
45 Postmus PE, Kerr KM, Oudkerk M, et al. Early and locally advanced
non-small-cell lung cancer (NSCLC): ESMO clinical practice
guidelines for diagnosis, treatment and follow-up. Ann Oncol 2017;
28 (suppl 4): iv1–iv21.
46 Pedersen JH, Rzyman W, Veronesi G, et al. Recommendations from
the European Society of Thoracic Surgeons (ESTS) regarding
computed tomography screening for lung cancer in Europe.
Eur J Cardiothorac Surg 2017; 51: 411–20.
47 Rampinelli C, De Marco P, Origgi D, et al. Exposure to low dose
computed tomography for lung cancer screening and risk of cancer:
secondary analysis of trial data and risk-benefit analysis. BMJ 2017;
356: j347.
48 Tammemagi MC, Berg CD, Riley TL, Cunningham CR, Taylor KL.
Impact of lung cancer screening results on smoking cessation.
J Natl Cancer Inst 2014; 106: dju084.
49 Ashraf H, Tonnesen P, Holst Pedersen J, Dirksen A, Thorsen H,
Dossing M. Effect of CT screening on smoking habits at 1-year
follow-up in the Danish Lung Cancer Screening Trial (DLCST).
Thorax 2009; 64: 388–92.
50 Brain K, Carter B, Lifford KJ, et al. Impact of low-dose CT screening
on smoking cessation among high-risk participants in the UK Lung
Cancer Screening Trial. Thorax 2017; 72: 912–18.
51 Fintelmann FJ, Bernheim A, Digumarthy SR, et al. The 10 pillars of
lung cancer screening: rationale and logistics of a lung cancer
screening program. Radiographics 2015; 35: 1893–908.

52 McKee BJ, Regis SM, McKee AB, Flacke S, Wald C. Performance of
ACR lung-RADS in a clinical CT lung screening program.
J Am Coll Radiol 2016; 13 (suppl 2): R25–R29.
53 Pinsky PF, Gierada DS, Black W, et al. Performance of lung-RADS
in the National Lung Screening Trial: a retrospective assessment.
Ann Intern Med 2015; 162: 485–91.
54 Mehta HJ, Mohammed TL, Jantz MA. The American College of
Radiology lung imaging reporting and data system: potential
drawbacks and need for revision. Chest 2017; 151: 539–43.
55 Li K, Yip R, Avila R, Henschke CI, Yankelevitz DF. Size and growth
assessment of pulmonary nodules: consequences of the rounding.
J Thorac Oncol 2016; 12: 657–62.
56 McWilliams A, Tammemagi MC, Mayo JR, et al. Probability of
cancer in pulmonary nodules detected on first screening CT.
N Engl J Med 2013; 369: 910–19.
57 van Riel SJ, Ciompi F, Jacobs C, et al. Malignancy risk estimation of
screen-detected nodules at baseline CT: comparison of the PanCan
model, Lung-RADS and NCCN guidelines. Eur Radiol 2017;
27: 4019–29.
58 Heuvelmans MA, Walter JE, Vliegenthart R, et al. Disagreement of
diameter and volume measurements for pulmonary nodule size
estimation in CT lung cancer screening. Thorax 2017; published
online Oct 22. DOI: 10.1136/thoraxjnl-2017-210770.
59 Heuvelmans M, Vliegenthart R, van Ooijen P, et al. Nodule size is
poorly represented by nodule diameter in low–dose CT lung cancer
screening. IASLC 17th world conference on lung cancer; Vienna,
Austria; Dec 4–7, 2016. P1.03–042.
60 Public Health England. Breast screening: programme consolidated
standards. London: Public Health England, 2017.
61 Perry N, Broeders M, de Wolf C, Tornberg S, Holland R,
von Karsa L. European guidelines for quality assurance in breast
cancer screening and diagnosis. Fourth edition—summary
document. Ann Oncol 2008; 19: 614–22.
62 LCRB Dutch breast cancer—focus on early disease. 2017.
http://www.lrcb.nl/en/breast-cancer/audit/ (accessed May 15, 2017).
63 CT Volumetry Technical Committee. Lung nodule assessment in
CT screening profile. 2017. QIBA. http://qibawiki.rsna.org/
images/f/fb/QIBA_CT_Vol_
LungNoduleAssessmentInCTScreening_2017.07.rev15.pdf (accessed
Aug 01, 2017).
64 Heuvelmans MA, Vliegenthart R, de Koning HJ, et al.
Quantification of growth patterns of screen-detected lung cancers:
the NELSON study. Lung Cancer 2017; 108: 48–54.
65 de Hoop B, Gietema H, van Ginneken B, Zanen P, Groenewegen G,
Prokop M. A comparison of six software packages for evaluation of
solid lung nodules using semi-automated volumetry: what is the
minimum increase in size to detect growth in repeated CT
examinations. Eur Radiol 2009; 19: 800–08.
66 Scholten ET, de Jong PA, Jacobs C, et al. Interscan variation of
semi-automated volumetry of subsolid pulmonary nodules.
Eur Radiol 2015; 25: 1040–47.
67 Chung K, Jacobs C, Scholten ET, et al. Lung-RADS category 4X:
does it improve prediction of malignancy in subsolid nodules?
Radiology 2017; 284: 264–71.
68 Midthun DE, Jett JR. Screening for lung cancer: the US studies.
J Surg Oncol 2013; 108: 275–79.
69 Bach PB, Mirkin JN, Oliver TK, et al. Benefits and harms of CT
screening for lung cancer: a systematic review. JAMA 2012;
307: 2418–29.
70 Callister ME, Baldwin DR, Akram AR, et al. British Thoracic Society
guidelines for the investigation and management of pulmonary
nodules. Thorax 2015; 70 (suppl 2): ii1–ii54.
71 Henschke CI, Naidich DP, Yankelevitz DF, et al. Early lung cancer
action project: initial findings on repeat screenings. Cancer 2001;
92: 153–59.
72 Henschke CI, Yankelevitz DF, Libby DM, Pasmantier MW,
Smith JP, Miettinen OS. Survival of patients with stage I lung
cancer detected on CT screening. N Engl J Med 2006; 355: 1763–71.
www.thelancet.com/oncology Published online November 27, 2017 http://dx.doi.org/10.1016/S1470-2045(17)30861-6 13
Policy Review
73 Wilson DO, Weissfeld JL, Fuhrman CR, et al. The Pittsburgh Lung
Screening Study (PLuSS): outcomes within 3 years of a first
computed tomography scan. Am J Respir Crit Care Med 2008;
178: 956–61.
74 Swensen SJ, Jett JR, Sloan JA, et al. Screening for lung cancer with
low-dose spiral computed tomography. Am J Respir Crit Care Med
2002; 165: 508–13.
75 Henschke CI, Yankelevitz DF, Yip R, et al. Lung cancers diagnosed
at annual CT screening: volume doubling times. Radiology 2012;
263: 578–83.
76 Carter D, Vazquez M, Flieder DB, et al. Comparison of pathologic
findings of baseline and annual repeat cancers diagnosed on CT
screening. Lung Cancer 2007; 56: 193–99.
77 Xu DM, Yip R, Smith JP, Yankelevitz DF, Henschke CI,
Investigators IE. Retrospective review of lung cancers diagnosed in
annual rounds of CT screening. AJR Am J Roentgenol 2014;
203: 965–72.
78 Cancer Research UK. Pulmonary nodule risk. 2017. https://itunes.
apple.com/gb/app/pulmonary-nodule-risk/id1142255949?mt=8
(accessed May 15, 2017).
79 Herder GJ, van Tinteren H, Golding RP, et al. Clinical prediction
model to characterize pulmonary nodules: validation and added
value of 18F-fluorodeoxyglucose positron emission tomography.
Chest 2005; 128: 2490–96.
80 US Preventive Services Task Force. Final Update Summary: Lung
Cancer: Screening. 2015. http://www.uspreventiveservicestaskforce.
org/Page/Document/UpdateSummaryFinal/lung-cancer-screening
(accessed July 26, 2017).
81 Yousaf-Khan U, van der Aalst C, de Jong PA, et al. Final screening
round of the NELSON lung cancer screening trial: the effect of a
2.5-year screening interval. Thorax 2017; 72: 48–56.
82 Ten Haaf K, Tammemagi MC, Bondy SJ, et al. Performance and
cost-effectiveness of computed tomography lung cancer screening
scenarios in a population-based setting: a microsimulation
modeling analysis in Ontario, Canada. PLoS Med 2017; 14: e1002225.
83 Pastorino U, Rossi M, Rosato V, et al. Annual or biennial CT
screening versus observation in heavy smokers: 5-year results of the
MILD trial. Eur J Cancer Prev 2012; 21: 308–15.
84 Duffy SW, Field JK, Allgood PC, Seigneurin A. Translation of
research results to simple estimates of the likely effect of a lung
cancer screening programme in the United Kingdom. Br J Cancer
2014; 110: 1834–40.
85 Yankelevitz D, Henschke C. Lung cancer: low-dose CT screening—
determining the right interval. Nat Rev Clin Oncol 2016; 13: 533–34.
86 Yousaf-Khan U, van der Aalst C, de Jong PA, et al. Risk stratification
based on screening history: the NELSON lung cancer screening
study. Thorax 2017; 72: 819–24.
87 Patz EF Jr, Greco E, Gatsonis C, Pinsky P, Kramer BS, Aberle DR.
Lung cancer incidence and mortality in National Lung Screening
Trial participants who underwent low-dose CT prevalence
screening: a retrospective cohort analysis of a randomised,
multicentre, diagnostic screening trial. Lancet Oncol 2016;
17: 590–99.
88 Dunn CE, Edwards A, Carter B, Field JK, Brain K, Lifford KJ.
The role of screening expectations in modifying short-term
psychological responses to low-dose computed tomography lung
cancer screening among high-risk individuals. Patient Educ Couns
2017; 100: 1572–79.
89 Field JK, Duffy SW, Devaraj A, Baldwin DR. Implementation
planning for lung cancer screening: five major challenges.
Lancet Respir Med 2016; 4: 685–87.
90 Ciompi F, Chung K, van Riel SJ, et al. Towards automatic
pulmonary nodule management in lung cancer screening with
deep learning. Sci Rep 2017; 7: 46479.
91 Field JK. Perspective: the screening imperative. Nature 2014; 513: S7.