Effectiveness of computed tomography screening for lung cancer: systematic reviews

The following are executive summaries of monographs published recently by the NHS R&D Health Technology Assessment programme, which publishes around 50 such reports a year. All the executive summaries and full reports are available at www.ncchta.org where they may be viewed and downloaded free of charge.

Printed full versions of all monographs can be obtained at a cost of £20 each from: York Publishing Services Ltd, PO Box 642, York, YO31 7WX. Tel: 0870 1616662; Fax: 0870 1616663; E-mail: ncchta@yps-publishing.co.uk. A fully searchable CD-ROM containing all the HTA monographs is also available free of charge from the above address.

Effectiveness of computed tomography screening for lung cancer: systematic reviewsC. Black, A. Bagust, A. Boland, S. Walker, C. McLeod, R. De Verteuil, J. Ayres, L. Bain, S. Thomas, D. Godde and N. Waugh

Background

Screening for lung cancer has been the subject of debate for the past three decades. This has largely stemmed from the results of chest X-ray screening studies where improvements in survival were obtained but without reductions in disease-specific, or total, mortality. The debate raises two issues: the design of studies to evaluate screening for lung cancer, in particular the choice of comparator; and the potential role of over-diagnosis of well-differentiated,slow-growing tumours that would not have led to symptoms or death in the lifetime of the affected patient.

Lung cancer is the leading cause of death from cancer in the UK, killing approximately 34,000 people per year. By the time symptoms develop, the tumour is often at an advanced stage and the prognosis is bleak. Treatment at a less advanced stage of disease with surgical resection has been shown to substantially reduce mortality. Screening would be attractive if it could detect presymptomatic lung cancer at a stage when surgical intervention is feasible.

Objectives

The aim of this review is to examine the clinical and cost-effectiveness of screening for lung cancer using computed tomography (CT) to assist policy making and to clarify research needs.

Methods

Search strategy. Fifteen electronic databases and Internet resources were searched from 1994 until December 2004/January 2005. In addition,bibliographies of the retrieved articles were searched and the register of projects held by the International Network of Agencies for HTA (INAHTA) was also checked.

Inclusion/exclusion criteria

Studies were included where screening for lung cancer was the principal theme of the paper. The initial search was for randomised trials in which survival in a group receiving CT screening was compared with a group not screened, but because of the lack of such studies, no restriction was placed on study type. Studies were reviewed by two authors independently.

Data extraction

Data extraction included details of the screening protocol, follow-up,diagnosis and participants. Information was sought about test characteristics,including sensitivity and specificity. The checklists and methods described in NHS Centre for Reviews and Dissemination (CRD) Report 4 were used for the quality assessment of studies.

Analysis

Separate narrative summaries were performed for the clinical effectiveness and cost-effectiveness. Cost-effectiveness analysis resulting in a cost per quality-adjusted life-year was not feasible, therefore the main elements of such an appraisal were summarised and the key issues relating to the existing evidence base were discussed.

Results

Summary of clinical effectiveness. In total, 12 studies of CT screening for lung cancer were identified, including two randomised controlled trials (RCTs) and ten studies of screening without comparator groups. The quality of reporting of these studies was variable, but the overall quality was adequate. The two RCTs were of short duration (1 year) and therefore there was currently no evidence that screening improves survival or reduces mortality. The proportion of people with abnormal CT findings varied widely between studies(5-51 per cent). The prevalence of lung cancer detected was between 0.4 and 3.2 per cent (number need to screen to detect one lung cancer=31-249). Incidence rates of lung cancer were lower (0.1-1 per cent per year). Detection of stage I and resectable tumours was high, 100 per cent in some studies. Adverse events,as a result of investigation or surgery, or the screening process per se were poorly reported. Incidental findings of other abnormalities requiring medical follow-up were reported to be as high as 49 per cent.

Summary of cost-effectiveness

Six full economic evaluations of population CT screening programmes for lung cancer were included in the review. The magnitude of cost-effectiveness ratios reported vary widely. None was set in the UK and generalisation was complicated by wide variation in the data used in different countries and a paucity of UK data for comparison. All six made the fundamental assumption that screening with CT for lung cancer reduced mortality. At the current time, there is no evidence to support that assumption.

Economic appraisal

In the absence of evidence of health gains from screening for lung cancer, in terms of either quantity or quality of life, and faced with a range of uncertainties, from the frequency of abnormal screening findings within a population to the natural history of screening detected lung cancers, it is not feasible at the current time to develop accurately and meaningfully an economic argument for CT screening for lung cancer in the UK. For subgroups, in particular certain occupational groups, there is evidence of increased risk of lung cancer, but the role of screening has not been demonstrated by the current studies.

Conclusions

The accepted National Screening Committee criteria are not currently met,with no RCTs, no evidence to support clinical effectiveness and no evidence of cost-effectiveness.

Recommendations for research

In terms of what information is needed to assist decision-making about CT screening for lung cancer, the following research priorities were identified:

• RCT evidence is needed about the effect of CT screening on mortality, either with whole-population screening or for particular subgroups. One such trial is underway in the USA, recruiting 50,000 participants, and is due to end in 2009,although final follow-up will not complete until around 2014.

• UK data about the rate of positive screening with CT and detected lung cancers could be obtained from an RCT or a cohort study. Even relatively small-scale studies would provide valuable information when trying to assess the generalisability of RCT data currently being conducted elsewhere.

• There is a need to understand better the natural history and epidemiology of screening-detected lung cancers, particularly small, well-differentiated adenocarcinomas. This could be met, in part, by lung cancer screening RCTs or cohort studies, but a review of existing published epidemiological and pathological data, along with primary analysis of UK lung cancer epidemiology,would usefully inform current understanding.

• Information about the quality of life impact of CT screening, acceptability of screening, and uptake and retention rates in the UK would be valuable in any future assessment of the cost-effectiveness of screening in the UK.

• Increased collection is needed of UK health service data regarding resource use and safety data for lung cancer management and services.

• Research is needed into the feasibility and logistics of tracing people who have in the past worked in industry where there was exposure to lung carcinogens.

©2006 Crown Copyright

(C. Black, R. De Verteuil, L. Bain, S. Thomas and N. Waugh are based at the Department of Public Health, University of Aberdeen, UK. A. Bagust is based at the University of Liverpool Management School, UK. A. Boland and C. McLeod are based at the Liverpool Reviews and Implementation Group (LRiG), University of Liverpool, UK. S. Walker is based at the Department of Radiology, Aberdeen Royal Infirmary, UK. J. Ayres is based at the School of Medicine, Environmental and Occupational Medicine, University of Aberdeen, UK. D. Godden is based at the School of Medicine, University of Aberdeen, UK.)