Elsevier

Lung Cancer

Volume 161, November 2021, Pages 136-140
Lung Cancer

Analysis of the baseline performance of five UK lung cancer screening programmes

https://doi.org/10.1016/j.lungcan.2021.09.012Get rights and content

Highlights

  • Minimising harms is critical in lung cancer screening implementation.

  • We report cumulative real-world lung cancer screening data from the UK.

  • Reported harms, including false positive and benign resection rates, are low.

  • Outcomes are comparable to, and in some aspects superior to, published RCTs.

Abstract

Introduction

Low-dose CT (LDCT) screening reduces lung cancer specific mortality. Several countries, including the UK, are evaluating the clinical impact and cost-effectiveness of LDCT screening using the latest evidence. In this paper we report baseline screening performance from five UK-based lung cancer screening programmes.

Methods

Data was collected at baseline from each screening programme. Measures of performance included prevalence of screen detected lung cancer, rate of surveillance imaging for indeterminate findings and surgical resection rates. Screening related harms were assessed by measuring false positive rates, number of invasive tests with associated complications in individuals without lung cancer and benign surgical resection rates.

Results

A total of 11,148 individuals had a baseline LDCT scan during the period of analysis (2011 to 2020). Overall, 84.7% (n = 9,440) of baseline LDCT scans were categorised as negative, 11.1% (n = 1,239) as indeterminate and 4.2% (n = 469) as positive. The prevalence of screen detected lung cancer was 2.2%, ranging between 1.8% and 4.4% for individual programmes. The surgical resection rate was 66% (range 46% to 83%) and post-surgical 90-day mortality for those with lung cancer 1.2% (n = 2/165). The false positive rate was 2% (n = 219/10,898) and of those with a positive result, one in two had lung cancer diagnosed (53.3%). An invasive test was required in 0.6% (n = 61/10,898) of screening attendees without lung cancer; there were no associated major complications or deaths. The benign surgical resection rate was 4.6% (n = 8/173), equating to 0.07% of the screened population.

Discussion

The performance of UK-based lung cancer screening programmes, delivered within or aligned to the National Health Service, compares favourably to published clinical trial data. Reported harms, including false positive and benign surgical resection rates are low. Ongoing monitoring of screening performance is vital to ensure standards are maintained and harms minimised.

Introduction

Lung cancer is the world’s leading cause of cancer related death [1]. In the United Kingdom (UK) there are over 45,000 new cases and 35,000 deaths each year [2]. The symptomatic presentation of lung cancer is characteristically associated with advanced disease when treatments are ineffective and survival is very poor. Screening high-risk smokers and former smokers with LDCT detects early stage disease, prior to the development of symptoms, and reduces lung cancer specific mortality. This has been definitively demonstrated in two large randomised controlled trials (RCT), the National Lung Screening Trial (NLST) and the NELSON trial [3], [4]. Screening with LDCT is being implemented in the United States, Canada, some East Asian countries and is being considered across Europe [5]. In the UK, a number of implementation pilots [6], [7], [8] and research studies [9], [10], [11] have demonstrated that screening can be successfully delivered within or aligned to the National Health Service (NHS). Building on these solid foundations NHS England recently announced funding (£71 million) to undertake an expanded screening programme at a number of sites across the country [12].

A critically important requirement for screening implementation is the minimisation of harms. It is also important for individuals who are considering the offer of screening to support informed decision-making by ensuring benefits and risks are clearly communicated. Harms associated with lung cancer screening are well described. These include overdiagnosis, anxiety or psychological harm related to screening outcomes, radiation exposure, false negatives and the investigation and treatment of false positive findings [5]. Although well described, the extent of harm from lung screening is variably reported. Determining the prevalence and clinical significance of screening related harms, for instance due to investigation or treatment of benign disease, is a key measure of screening quality [5].

Results from large RCTs are commonly used as a benchmark for screening performance and to quantify the potential risks and benefits. However, data extrapolated from RCTs may not provide an accurate representation of screening performance in a ‘real world’ setting. This may be due to important differences in population selection, healthcare setting or pulmonary nodule management protocols. One recent study demonstrated how modern pulmonary nodule management may have reduced harm in the NLST [13]. In this paper we present cumulative data from five UK-based lung cancer screening programmes; focusing on performance and measures that reflect possible screening related harms. This is the first-time screening harms data have been examined in detail within the NHS and out with the confines of a single research trial. We present this data as infographics which, after careful evaluation with the target population first, could be used to communicate the harms of screening to those considering screening participation.

Section snippets

Methods

Aggregate data from five UK lung cancer screening programmes is included in the analysis. This includes two RCTs, the UK Lung Cancer Screening Trial (UKLS) [9] and the Lung Screen Uptake Trial (LSUT) [10], an observational cohort study, the Nottingham Lung Health MOT [14], and two NHS commissioned screening services: the Manchester Lung Health Check pilot [6], [7] and the Liverpool Healthy Lung Project [8]. All sites performed a baseline screening round, Manchester was the only site to perform

Results

A brief overview of each screening programme is provided in Table 1. A total of 11,148 individuals had a baseline LDCT scan and are included in the analysis; ranging from 361 to 6,639 at each site. Overall, 84.7% (n = 9,440/11,148) were categorised as negative and either returned to annual screening or discharged depending on individual site protocols. 11.1% (n = 1,239/11,148) had an indeterminate screening outcome and required interval imaging. The proportion classified as positive, including

Discussion

In this paper we report baseline screening performance from five independent UK-based lung cancer screening programmes including over 11,000 participants. The overall prevalence of lung cancer was 2.2%. This compares to 0.9% and 1% in NELSON and NLST respectively. The higher prevalence observed across the UK programmes may reflect the use of individual risk-based selection criteria and the location of services in areas of high socio-economic deprivation. Only 4.2% of participants had a positive

CRediT authorship contribution statement

Haval Balata: Conceptualization, Data curation, Investigation, Methodology, Validation, Writing – original draft. Mamta Ruparel: Data curation, Investigation, Validation, Writing – review & editing. Emma O’Dowd: Data curation, Investigation, Validation, Writing – review & editing. Martin Ledson: Data curation, Investigation, Validation, Writing – review & editing. John K. Field: Investigation, Writing – review & editing. Stephen W. Duffy: Writing – review & editing. Samantha L. Quaife: Writing

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (19)

There are more references available in the full text version of this article.

Cited by (30)

  • Lung cancer screening

    2023, The Lancet
  • Early Diagnosis and Lung Cancer Screening

    2022, Clinical Oncology
    Citation Excerpt :

    In comparison, NELSON, where volumetric measurements, larger positive size definitions and surveillance groups identified, reported much lower false-positive rates of 1.2%. More recently, real-world data from five UK-based screening programmes reported an overall false-positive rate of 2.2%, significantly lower than that reported in NLST and the majority of European screening trials outside of NELSON [71]. The ever-increasing numbers of screening programmes internationally mean that incidental findings are becoming increasingly prevalent internationally.

View all citing articles on Scopus
View full text