Gabrijela Kocjan, N Magee, M Novelli, Mej Callister, K Sridharan, M.D. Peake, J Allen, Stephen G. Spiro, Penny Shaw, V Ashford-Turner, Robert C. Rintoul, Mary Falzon, Sam M. Janes, P Dhillon, N Counsell, Richard Booton, Jeremy George, Nyasha Chinyanganya, C Griffths, Pallav L. Shah, and A. Hackshaw
Screening for the early detection of lung cancers should increase the percentage of operable tumours, thus improving cure rates. A large randomised US trial showed that CT screening moderate/heavy smokers is effective but expensive, with a high false-positive rate. We designed LungSEARCH in 2006 to target screening in higher-risk subjects. Because most tumours in the UK were of squamous-histology, we hypothesised that sputum cytology plus cytometry would be an effective initial screen, only offering more intensive/expensive tests to those with abnormal sputum. Eligibility criteria were: current/former smokers (≥20 pack-years and/or smoked ≥20 years), GOLD-defined COPD, no prior cancer. Subjects were randomised to surveillance or a control group, and each followed for 5 years. Screened subjects provided sputum for central assessment, and those with abnormal results (cytology: low/high-grade squamous intraepithelial lesions, and/or cytometry: abnormal ploidy) were referred for annual low-dose CT and autofluorescence bronchoscopy (AFB) for the remainder of the trial, with diagnostic investigations when cancer suspected by abnormal CT/AFB. Sputum-negatives provided annual sputum samples only. Control subjects had a chest X-ray when they reached 5 years. Primary objective: to show a higher proportion of early stage cancers using surveillance than controls. 1568 subjects were recruited (target 1300) from GPs or chest clinics around 10 UK centres (August 2007–March 2011): 785 screened, 783 controls. Mean age 63 years; males 52%; current (56%), former (44%) smokers; mild (25%), moderate (75%) COPD; from GPs (79%). >90% screened subjects provided sputum samples in their first year. After 5 years, the overall sputum-positive rate is 33%; 30% (236/785) had a CT scan and 25% (193/785) had an AFB at any time. Of those who had a CT scan 19% (45/236) were abnormal (lung nodule(s) ≥9 mm); and of those who had AFB 3% (5/193) had severe dysplasia or worse. 79 lung cancers have been identified to date via the centres/national registry: 43 surveillance and 36 control. But awaiting staging details for 6 surveillance and 14 control cases. Preliminary results are promising: 57% (surveillance) versus 41% (controls) of cancers were diagnosed with stage I/II non-small-cell-lung cancer or limited disease small-cell-lung cancer. Final data available later in 2016.