From incidental findings and screening for chronic obstructive pulmonary disease (COPD) to surveillance imaging protocols and the advent of artificial intelligence (AI), the authors of a new meta-analysis examine insights and emerging trends from the last two decades of research on the use of low-dose computed tomography (CT) in lung cancer screening.
When should surveillance chest computed tomography (CT) be discontinued for lung cancer survivors? Are incidental findings on low-dose CT over-diagnosed at academic institutions? Should chronic obstructive pulmonary disease (COPD) be more of a consideration when assessing low-dose chest CT scans? Can artificial intelligence have an impact in improving lung nodule detection rates on low-dose CT?
In a recently published meta-analysis in the Lancet, researchers considered these questions and more in their review of clinical trials, guidelines, systematic reviews, and other original research on lung cancer screening with low-dose CT published between 2000 and April 2022. Here are nine key takeaways.
1) There doesn’t appear to be a consensus approach in regard to the timing for discontinuing surveillance imaging in lung cancer survivors. Pointing to registry data from England and Germany, the meta-analysis authors noted that lung cancer survivors have an elevated risk for a second primary lung cancer 10 years after the initial diagnosis as well as other smoking-related second primary cancers involving laryngeal cancer, head and neck cancer and esophageal squamous cell carcinoma. While current guidelines recommend chest CT every six months for two to three years after completed lung cancer treatment and subsequent annual chest CT exams, the meta-analysis authors said many guidelines do not specify an endpoint for surveillance CT imaging in this patient population.
2) Lung cancer accounts for 25 percent of second primary malignancies in patients with a history of bladder cancer.
3) Researchers noted incidental findings on 67 percent of low-dose CT scans obtained in university imaging centers and 28 percent of low-dose CT exams from community imaging centers.
4) Only 15 percent of the aforementioned incidental findings required follow-up assessment, according to the meta-analysis authors.
5) Noting the ability of deep learning-enabled image reconstruction techniques to reduce noise and radiation dosing, the researchers also pointed out that artificial intelligence (AI) algorithms have shown sensitivity rates ranging between 83 to 97 percent and specificity rates ranging from 82 to 98 percent for the detection of lung nodules.
“Artificial intelligence for lung nodule classification could (optimize) nodule management by reducing unnecessary workup of benign nodules, reducing time to diagnosis of malignant nodules, and reducing inter-reader variability. In the future, combing artificial intelligence with existing guidelines such as Lung-RADS could provide an improved framework for nodule management,” wrote Florian J. Fintelmann, M.D., the head of Thoracic Imaging Percutaneous Thermal Ablation at Massachusetts General Hospital and an associate professor at Harvard Medical School, and colleagues.
6) In one study with health equity ramifications, researchers noted that 23 percent of people who participated in mobile CT lung screening program for a second time said they would be less likely to participate in a hospital-based lung screening program.
7) Depending upon the screening criteria employed, studies noted low-dose CT detection of mild emphysema that ranged between 24 to 63 percent of lung cancer screening participants.
8) Emphasizing that low-dose CT “is an opportunity to diagnose” chronic obstructive pulmonary disease (COPD), the meta-analysis authors suggested that clinicians go beyond age and pack-year criteria in the consideration of potential COPD.
“Combined screening criteria based on age and smoking history miss an important proportion of individuals who will suffer from lung cancer, in part due to suboptimal consideration of COPD,” noted Fintelmann and colleagues.
9) Emphasizing that extensive coronary calcium is a strong predictor of cardiovascular events, the meta-analysis authors noted that two consensus documents have advocated for coronary calcium reporting to be standard for chest CT exams (including the use of low-dose CT for lung cancer screening).