P62.09 A Prospective Cohort Evaluation of the Sensitivity and Specificity of the Chest X-Ray for the Detection of Lung Cancer in Symptomatic Adults

Abstract

The chest radiograph (CXR) plays a pivotal role in the triage of symptomatic individuals with potential lung malignancy. In the United Kingdom the National Institute of Clinical Excellence recommends CXR as an initial investigation in those aged over 40 with specified symptoms or in the presence of one symptom and smoking history. Urgent specialist evaluation is advised if the CXR demonstrates an abnormality suggestive of malignancy. The evidence regarding the effectiveness of the CXR in this role is incomplete and based upon the retrospective evaluation of populations identified with lung cancer. The design of these studies prevents the evaluation of correctly excluded cases and an estimation of the test specificity. This prospective cohort study aimed to establish the accuracy of the CXR in the investigation of malignancy amongst symptomatic adults over the age of 50 for intrathoracic malignancies Secondary aims were to evaluate how a history of smoking alters the test characteristics, the role of CXR follow-up, and whether the reported presence of abnormalities for which no follow-up is advised predict the presence of malignancy. The study population was a cohort of consecutive “self-request” CXR studies prospectively collated between January 2011 and October 2016 as part of the National Awareness and Early Diagnosis Initiative. The initiative combined an early lung cancer diagnosis awareness campaign with a policy of open access to a CXR. Individuals were able to obtain a CXR providing they met eligibility criteria. The criteria were symptoms persisting for at least 3 weeks (cough, fatigue, shortness of breath, chest pain, loss of appetite or loss of weight) and age over 50. Patient’s self-reported paper questionnaires collected during the study period formed the cohort for this study, with patient subsequently cross referenced with the regional cancer registry to identify intrathoracic malignancies. The CXR report was retrieved and coded using a novel definition of a positive test as a CXR resulting in investigation with CT. Studies were positive if the report documented an abnormality with a recommendation which directly (or indirectly after a CXR or clinical follow-up) led to investigation with CT. Studies which did not result in investigation with CT were considered negative. 8,948 CXR outcomes were evaluated. 496 positive studies led to a diagnosis of 80 patients with Non-Small Cell Lung Cancer (NSCLC) amongst 101 primary intrathoracic malignancies. Within two-years, a cumulative total of 133 NSCLC amongst 168 primary intrathoracic malignancies were observed. The sensitivity and specificity for NSCLC were 76% (95%CI 68-84) and 95% (95%CI 95-96) within 1-year and 60% (95%CI 52-69) and 95% (95%CI 95-96) within 2-years. The 2-yr positive and negative likelihood ratios were 12.8 and 0.4. The results did not differ for NSCLC compared to all primary malignancies. A positive test strongly increases the probability of malignancy whereas a negative test moderately reduces the risk. A quarter of NSCLC will have a negative CXR in the year prior to diagnosis. The findings allow the risk of malignancy following a negative test to be estimated.