The British Journal of Radiology, 82 (2009), 529531

COMMENTARY

CT lung cancer screening in the UK

A J EDEY,

MRCP, FRCR

and D M HANSELL,

MD, FRCP, FRCR

Department of Radiology, Royal Brompton Hospital, Sydney Street, London SW3 6NP, UK

ABSTRACT. Lung cancer is the most common cause of cancer-related death in the UK.
Despite aggressive primary prevention measures and improved medical care, the 5-year
survival rate is less than 10% for patients in the UK who present with symptoms. The
possibility of CT screening for lung cancer provides some hope of reducing mortality.
However, the case for screening remains unproven. This article explores the issues
surrounding lung cancer screening in the context of historical studies, trials in progress
and tentative plans for a UK CT lung cancer screening trial.

Lung cancer is the most common cause of cancerrelated death in the UK. Whilst the attraction of screening to reduce this appalling rate of mortality is obvious,
its value in the specific context of lung cancer is highly
controversial. In the absence of screening, the
Department of Healths National Cancer Plan [1] and
recent Cancer Reform Strategy [2] set out a framework to
ensure delivery of consistent care across the NHS for
patients who present clinically with lung cancer.
However, despite these measures and advances in
surgical and oncological techniques, the onset of symptoms still signals a dismal outlook, with a ,10% survival
rate at 5 years in the UK [3]. Additional UK governmental legislation to curb smoking (an underlying risk
factor in up to 90% of cases of lung cancer) has
contributed to the ongoing decline in the size of the
UK smoking population. However, there remains a large
cohort of smokers and, importantly, ex-smokers who are
at risk of developing lung cancer. It is this high-risk
population for whom CT screening may prove valuable
in preventing premature death. Some of the issues
surrounding CT screening for lung cancer, in the context
of historical and ongoing trials, are now reviewed.
Early attempts at lung cancer screening using chest
radiography were abandoned in the 1980s as a result of
several large trials that failed to show any mortality
benefit [4]. More recently, advocates of CT screening
have highlighted its clear superiority compared with
chest radiographs for detecting small lung cancers and
the applicability of low radiation dose protocols. As a
result, a number of observational studies were carried
out in the 1990s with Henschke and colleagues in the
vanguard. These studies provided several valuable
insights. Firstly, it was confirmed that low-dose CT
detects more lung cancers than do chest radiographs [5].
Secondly, low-dose CT is highly effective at detecting
early-stage lung cancers (between 50% and 91% of
screening-detected cancers are Stage 1a [6]). Thirdly,
Address correspondence to: D M Hansell, Department of
Radiology, Royal Brompton Hospital, Sydney Street, London SW3
6NP, UK. E-mail: [email protected]

The British Journal of Radiology, July 2009

Received 16 January 2009

Revised 13 March 2009
Accepted 24 March 2009
DOI: 10.1259/bjr/17503608
2009 The British Institute of
Radiology

resection of Stage 1 cancers prolongs survival times: in

the Early Lung Cancer Action Project (ELCAP), the 10year survival rate for resected Stage 1 tumours was 92%
[5], in contrast with an average 5-year survival rate of
62% for Stage 1 cancer reported outside the screening
literature [7]. This last point is the most contentious, with
many experts pointing out that an improved 5-year
survival rate is not equivalent to a reduction in
population mortality (the ultimate point of any screening
programme). Although the ELCAP authors found that
all patients with untreated Stage 1a cancer died within
5 years of detection [8, 9], a different study by Bach et al
[10] found that no population mortality benefit was
evident after 5 years of screening (the screened population was compared with a validated mathematical
population model as a control) [10]. However,
Henschke and Yankelevitz [5] made the obvious argument that a true mortality benefit will only become
apparent after longer periods of follow-up. Furthermore,
they pointed out that, in the early years of a screening
programme, asymptomatic individuals with relatively
late-stage cancers will be detected who may not benefit
from treatment. Thus, deaths in such patients will
effectively mask the benefits of screening. Theoretically,
once screening is established, new tumours will be
detected at an earlier stage and will therefore be
amenable to intervention. There are concerns about the
clinical relevance of small screening-detected cancers, as
a significantly higher proportion (up to 70% [10]) are
detected than would be expected to become clinically
evident within the lifespan of a matched non-screened
population. The apparent discrepancy between the
prevalence of screening-detected disease and expected
population prevalence of lung cancer has caused some to
question an underlying assumption of screening: namely
that early-stage screening-detected cancer presages the
inexorable development of end-stage disease which, if
detected and treated, will prevent lung cancer-related
deaths. Instead, a bipartite model for lung cancer has
been proposed [10, 11]. In this model, some cancers are
indolent and non-fatal, whereas others develop rapidly
and elude screening, presenting as incurable interval
529

A J Edey and D M Hansell

cancers (this dichotomy is analogous to prostate cancer).

Without data from randomised controlled trials, these
issues (as well as those regarding the known biases of
screening) remain unresolved and population CT screening for lung cancer is not recommended at present in the
UK [12].
In order to formulate a strategy to investigate lung
cancer screening more efficiently, an international consortium with expertise in lung cancer screening was
formed in 2001. The European Union/United States
Spiral CT Collaborative Group sought to establish
common practices that would allow data from future
trials to be pooled easily [13]. The advantages of
international collaboration are twofold. First, the larger
the data set available, the shorter the trial period
required to show a significant reduction (or otherwise)
in mortality. This would provide data over a shorter time
frame and address the fundamental question about the
efficacy of CT screening for lung cancer. Second, mass
population-based screening trials are expensive. Pooling
data allows countries with smaller research budgets, and
indeed smaller numbers of appropriate subjects, to
participate and contribute to the accumulation of data
while acquiring local screening expertise.
Currently, seven randomised controlled trials of lowdose CT screening are underway, involving approximately 82 000 individuals. These trials can be divided
into two types: (i) those in which the control (no CT) arm
employs radiographic screening and (ii) those with no
screening of any sort in the control arm. The former
include DepiScan [14], the Lung Screening Study [15]
and the National Lung Cancer Screening Trial (NLST)
[16]. The largest of these, the US NLST, was set up in
2002 and completed the recruitment of 50 000 smokers
and ex-smokers in 2004. It aims to answer the question of
whether CT is a better screening tool than chest radiography and has been calculated to have a power of 90%
to detect a 20% reduction in mortality. Results are
expected in 2012. The decision to use chest radiographs
in the control population was taken with a view to
integrating data from the Prostate, Lung, Colorectal and
Ovarian screening trial (which compares screening with
chest radiographs with usual care, i.e. no screening), in
effect forming a three-arm study. However, the obvious
criticism that NLST does not directly evaluate
whether CT screening is better than no screening
means that this question will remain unanswered. In
addition, NLST will initially provide only a 3-year
snapshot of changes in mortality, which may be
insufficient to detect true differences. The second group
of trials (mostly European) has an unscreened control
arm (or in the case of DANTE (Lung Cancer Screening
with Low Dose Spiral Computed Tomography) a single
baseline radiograph but no further chest radiographs at
follow-up) and includes DANTE [17], ITALUNG (The
Italian Lung Screening Trial) [18] and NELSON (The
Dutch Belgian Randomised Lung Cancer Screening Trial)
[19]. The NELSON trials, based in Belgium and
Denmark, comprise 16 000 and 4000 participants, respectively, and the data from the two centres will be pooled.
These trials aim to follow up over a 10-year period
(results are expected in 2014) and have mortality as their
primary outcome measure. The study will have 80%
power to detect a mortality reduction of 25% at 10 years
530

(with a level of significance set at 0.05). However, the

authors have raised concerns about the power calculations: if screening only reduces mortality by 20% then
they will require 40% more participants than currently
recruited to demonstrate an effect. In addition, variations
in recruitment and selection between the two groups
mean that there are theoretical differences between the
study populations (although both groups are at high risk
of lung cancer), which may influence the outcome.
Nevertheless, to date this is the largest randomised
controlled trial looking specifically at lung cancer
mortality and, as a consequence, it is being monitored
with much interest.
The UK has been relatively inactive in CT lung cancer
screening research. Although somewhat late in the day,
it is possible that a national multicentre UK study will
start, which, if successful, could be rolled out across the
country. Initiation of such a trial would not only add to
the sum of evidence but would also have the secondary
benefit of increasing expertise in the management of
incidentally detected pulmonary nodules. As yet, a
precise protocol for screening has not been developed
fully. Most of the ongoing randomised controlled trials
provide annual screening over a 3- or 4-year period for
individuals with negative baseline studies. Inevitably, in
a nationalised health service, cost is a significant, if not
over-riding, factor in developing a screening programme. Based on a one-off screening CT, the potential
incremental cost-effectiveness ratio for screening has
been calculated to be approximately 14 000 per qualityadjusted life year [20], which is within the limits set by
the National Institute for Clinical Excellence. However,
there are no data directly applicable to the UK that
assesses other screening models, e.g. with repeated CTs
at 2 years and 4 years, which is important as the time
between screening studies correlates with the incidence
of interval cancers [21]. In addition, the implications for
service provision, in terms of manpower and equipment,
are difficult to calculate and the estimates fraught.
Despite this, comparisons with breast screening highlight the massive requirements of implementing a
screening programme, and it is salutary to note that
there is still a shortfall in the number of trained clinicians
(particularly radiologists) 20 years after the UK programme was started.
Careful evaluation of the inherent risks from screening
(i.e. ionising radiation) and subsequent intervention for
true- and false-positive results are necessary if screening
is to be considered. In order to optimise the risk-tobenefit ratio, careful patient selection is required.
Selection of high-risk patients helps to increase the
specificity of screening, while excluding patients who
would not be fit for intervention both minimises risk and
reduces the expense of a screening programme. The
development of a validated lung cancer risk prediction
model based on a UK population by Field and colleagues
may provide a means of rationalising patient selection
for screening [22, 23]. Furthermore, a well-designed
follow-up process for positive screening studies can also
minimise unnecessary intervention. Thus, although 40%
of baseline CTs were positive in a recent observational
screening study [24], 96% of these were shown to be
benign on follow-up and did not require invasive
diagnostic tests. Significantly, of the patients who
The British Journal of Radiology, July 2009

Commentary: CT lung cancer screening in the UK

underwent surgery but had non-malignant disease, 90%

had bypassed the study protocol and would not have
been recommended for surgery as part of this US trial
[24]; it seems unlikely that, within the confines of the
NHS, UK patients would be able to circumvent trial
protocols quite so easily.
An inevitable consequence of screening by CT is the
detection of non-pulmonary abnormalities, e.g. coronary
artery calcification and thyroid nodules. Evidently lifethreatening disease cannot be ignored, but the risk of
intervening for otherwise unsuspected abnormalities,
in the context of a trial, is that it may obscure outcomes
and potentially alter all-cause mortality. However, the
frequency of incidental findings that have clinical
implications is reportedly low; in the NELSON study
[25], 1% of study participants were found to have
incidental abnormalities and, of this group, only one
patient had a non-pulmonary malignancy (which was
too advanced for clinical intervention). As a result, the
view of the NELSON investigators is against the
systematic searching for incidental findings on low-dose
screening CTs of the chest [25]. One advantage of this
approach is that it may allow trained technicians or
radiographers, working to a formal protocol, to act as a
first reader of screening studies, obviating the need to
scrutinise parts of the CT study other than the lungs.
At present, the role of CT screening for lung cancer is
uncertain. Although there are randomised controlled
trials underway around the world, the number of
individuals screened is relatively small, and more will
be needed to answer robustly the fundamental question
about the effect on mortality; international and national
collaboration would be one way of pursuing this goal.

References
1. Department of Health, UK. The NHS Cancer Plan: a plan for
investment a plan for reform. Available from: http://
www.dh.gov.uk/en/Publicationsandstatistics/Publications/
PublicationsPolicyAndGuidance/DH_4009609 [Accessed 17
December 2008].
2. Department of Health, UK. Cancer Reform Strategy. Available
from: http://www/dh.gov.uk/en/PublicationsandStatistics/
Publications/PublicationsPolicyAndGuidance/DH_081006
[Accessed 20 December 2008]
3. Coleman MP, Rachet B, Woods LM, Mitry E, Riga M,
Cooper N, et al. Trends and socioeconomic inequalities in
cancer survival in England and Wales up to 2001. Br J Cancer
2004;90:136773.
4. Ellis SM, Husband JE, Armstrong P, Hansell DM.
Computed tomography screening for lung cancer: back to
basics. Clin Radiol 2001;56:6919.
5. Henschke CI, Yankelevitz DF. CT screening for lung cancer:
update 2007. Oncologist 2008;13:6578.
6. Manser RL, Irving LB, Stone C, Byrnes G, Abramson M,
Campbell D. Screening for lung cancer. Cochrane Database
Syst Rev 2004;1:CD001991.
7. Mountain CF. Revisions in the international system for
staging lung cancer. Chest 1997;111:17107.
8. 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
26;355:176371.

The British Journal of Radiology, July 2009

9. Henschke CI. Clarifying enrollment procedures in the trial

of CT screening for lung cancer. N Engl J Med
2008;359:8713.
10. Bach PB, Jett JR, Pastorino U, Tockman MS, Swensen SJ,
Begg CB. Computed tomography screening and lung cancer
outcomes. JAMA 2007;297:95361.
11. Bach PB. Is our natural-history model of lung cancer
wrong? Lancet Oncol 2008;9:6937.
12. Black C, de VR, Walker S, Ayres J, Boland A, Bagust A, et al.
Population screening for lung cancer using computed
tomography, is there evidence of clinical effectiveness? A
systematic review of the literature. Thorax 2007;62:1318.
13. Field JK, Smith RA, Duffy SW, Berg CD, van KR, Henschke
CI, et al. The Liverpool Statement 2005: priorities for the
European Union/United States spiral computed tomography collaborative group. J Thorac Oncol 2006;1:4978.
14. Blanchon T, Brechot JM, Grenier PA, Ferretti GR, Lemarie E,
Milleron B, et al. Baseline results of the Depiscan study: a
French randomized pilot trial of lung cancer screening
comparing low dose CT scan (LDCT) and chest X-ray
(CXR). Lung Cancer 2007;58:508.
15. Gohagan JK, Marcus PM, Fagerstrom RM, Pinsky PF,
Kramer BS, Prorok PC, et al. Final results of the Lung
Screening Study, a randomized feasibility study of spiral
CT versus chest X-ray screening for lung cancer. Lung
Cancer 2005;47:915.
16. National Cancer Institute, USA. National Lung Cancer
Screening Trial. Available from: http://www.cancer.gov/
NLST [Accessed 17 December 2008].
17. Infante M, Lutman FR, Cavuto S, Brambilla G, Chiesa G,
Passera E, et al. Lung cancer screening with spiral CT:
baseline results of the randomized DANTE trial. Lung
Cancer 2008;59:35563.
18. Picozzi G, Paci E, Lopez PA, Bartolucci M, Roselli G, De FA,
et al. Screening of lung cancer with low dose spiral CT:
results of a three year pilot study and design of the
randomised controlled trial Italung-CT. Radiol Med
2005;109:1726.
19. van Iersel CA, de Koning HJ, Draisma G, Mali WP, Scholten
ET, Nackaerts K, et al. Risk-based selection from the general
population in a screening trial: selection criteria, recruitment and power for the Dutch-Belgian randomised lung
cancer multi-slice CT screening trial (NELSON). Int J
Cancer 2007 15;120:86874.
20. Whynes DK. Could CT screening for lung cancer ever be
cost effective in the United Kingdom? Cost Eff Resour Alloc
2008;6:5.
21. Tabar L, Faberberg G, Day NE, Holmberg L. What is the
optimum interval between mammographic screening
examinations? An analysis based on the latest results of
the Swedish two-county breast cancer screening trial.
Br J Cancer 1987;55:54751.
22. Cassidy A, Myles JP, van TM, Page RD, Liloglou T, Duffy
SW, et al. The LLP risk model: an individual risk prediction
model for lung cancer. Br J Cancer 2008;98:2706.
23. Field JK, Smith DL, Duffy S, Cassidy A. The Liverpool Lung
Project research protocol. Int J Oncol 2005;27:163345.
24. Wilson DO, Weissfeld JL, Fuhrman CR, Fisher SN, Balogh
P, Landreneau RJ, 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
1;178:95661.
25. van de Wiel JC, Wang Y, Xu DM, van der Zaag-Loonen HJ,
van der Jagt EJ, van Klaveren RJ, et al. Neglectable benefit
of searching for incidental findings in the Dutch-Belgian
lung cancer screening trial (NELSON) using low-dose
multidetector CT. Eur Radiol 2007;17:147482.

531