The Mortality Burden. Lung cancer remains the leading cause of cancer-related mortality worldwide. Despite advances in immunotherapy and targeted therapy that have modestly improved survival for advanced-stage patients, the prognosis at later stages has not dramatically improved. Early detection through screening offers the most powerful opportunity to shift outcomes.
Evidence for Screening. A 2022 Cochrane review of over 91,000 participants across eight randomized trials demonstrated that low-dose CT (LDCT) screening reduces lung cancer mortality by at least 21%, with an even greater benefit observed in women. The landmark NELSON and NLST trials confirmed that detecting lung cancer at earlier stages through LDCT leads to significant survival improvements.
European Expansion. Based on this compelling evidence, the European Council revised its cancer screening recommendations in 2022 to include lung cancer and urged European countries to assess feasibility of lung cancer screening (LCS) programs. The EU4Health SOLACE project and Europe's Beating Cancer Plan are actively facilitating rollout across Europe.
Stage Shift Evidence. Following the introduction of LDCT screening in the US, improved survival and a measurable shift toward stage I lung cancer detection have been documented. Catching lung cancer at stage I rather than stage III or IV translates to dramatically better survival -- making screening program design critically important.
Eligibility Criteria. Lung cancer screening is recommended for current and former smokers aged 50 to 75 years with at least 20 pack-years of smoking history. Former smokers who quit more than 10 to 15 years ago may be excluded from some protocols, though this exclusion is increasingly questioned given that aging itself increases lung cancer risk regardless of smoking recency.
Risk Model Refinement. Comprehensive risk prediction models that incorporate additional factors -- such as family history of lung cancer and occupational asbestos exposure -- have been developed to identify milder smokers who may still benefit from screening. Machine learning models incorporating clinical and biological parameters have recently been proposed and are awaiting validation.
Blood and Breath Biomarkers. Potential blood, exhaled air, or sputum biomarkers are being investigated to help identify early-stage lung cancer. While none have yet been validated for first-line patient selection, they could be combined with LDCT to increase specificity if their cost proves acceptable, serving as supplemental rather than replacement tools.
Screening Intervals. Annual LDCT has been the conventional approach, but a biennial (every two years) interval is increasingly proposed as a safe and more sustainable option after a negative scan. The European 4-IN THE LUNG RUN trial is investigating personalized invitation strategies and individualized screening frequencies based on risk profiles.
Scanner and Protocol Specifications. Effective LCS requires thin-slice volumetric CT scans acquired with a multidetector CT of 32 or more detector rows, a gantry rotation time of 0.5 seconds or less, and full chest coverage in under 10 seconds at full inspiration. Images should be reconstructed at slice thickness of 1.0 mm or less, preferably 0.75 mm or less.
Radiation Dose Targets. Keeping radiation exposure as low as possible is essential for a population-wide screening program. The effective dose target is below 1 mSv. Recommended CT dose index volumes are 0.4 mGy for patients under 50 kg, 0.8 mGy for patients 50-80 kg, and 1.6 mGy for patients over 80 kg. Automatic exposure control and organ dose modulation help achieve these targets.
The Role of Deep Learning Reconstruction. Iterative reconstruction or deep learning-based reconstruction algorithms must be used instead of conventional filtered back projection to reduce image noise at low radiation doses. Ultra-low-dose protocols approaching the dose of a standard chest X-ray have not yet been prospectively validated for LCS settings and cannot yet be recommended.
Consistency for Follow-up. Acquisition and reconstruction parameters must remain constant across serial examinations of the same patient. This consistency is critical for reliable volumetric follow-up of indeterminate lung nodules, since even small technical variations can affect apparent nodule volume measurements and growth calculations.
AI Capabilities in Screening. Artificial intelligence has the potential to significantly increase the efficiency of lung cancer screening programs. AI software tools have substantially improved over the past decade and can assist with nodule detection, volumetric measurement, characterization, and malignancy risk estimation. More than 10 CE-marked commercially available AI products exist in Europe for use as radiologist aids.
Detection and Sensitivity. AI nodule detection tools have the potential to increase sensitivity for pulmonary nodule identification and reduce reading time. This is particularly valuable in high-volume screening programs where radiologist fatigue and workload present real risks. Studies show these algorithms achieve comparable or slightly inferior performance compared to specialized radiologists.
Known Limitations. Current AI tools carry a risk of false negatives for pulmonary masses and mediastinal lesions. The scientific evidence on the independent performance of most commercially available products remains limited. Validation studies comparing AI as a second reader versus double reading by two radiologists are still needed.
The Best Use Case. At present, AI is best used as a second or concurrent reader rather than a replacement for trained radiologists. Visual reading by trained radiologists, complemented by thin-slab axial images and 10-15 mm maximum intensity projections (MIP), remains the primary reading approach with AI providing additional support and quality assurance.
Positive Screen Criteria for Solid Nodules. According to ESTI guidelines, solid nodules with a volume of 500 mm3 or greater at baseline represent a positive screening result and require referral to a multidisciplinary team for further diagnostic workup. Morphological features also trigger positive classification: spiculations, bubble-like lucencies, pleural indentation, complex cysts, and thick-walled cavitary nodules are all considered suspicious regardless of size.
Indeterminate Nodules and Follow-up Timing. Solid nodules between 100 and 250 mm3 require 6-month follow-up, with MDT referral if volume doubling time (VDT) is under 400 days. Those between 250 and 500 mm3 are re-evaluated at 3 months with MDT referral if VDT is under 250 days. At annual follow-up, a VDT under 500 days is considered suspicious. New nodules are usually inflammatory and 66% resolve on the next CT.
Negative Screen Results. Solid nodules smaller than 100 mm3 and nodules with clearly benign features represent a negative screening result. Intrapulmonary lymph nodes (IPLNs) -- which are small, subpleural, below the carina, and have sharp margins -- are very frequently the cause of apparently concerning nodules and should not trigger further investigation. Fat-containing nodules (hamartomas) are definitively benign when fat content is confirmed on mediastinum filter.
Subsolid Nodules. Pure ground glass and part-solid nodules (subsolid nodules, SSN) are considered negative if they measure less than 3 cm and have a solid component under 6 mm, unless suspicious morphology is present. Those with a solid component of 1 cm or larger at baseline require 1-month re-evaluation. Persistent SSN must develop a growing solid component or morphological changes to qualify as a positive result.
The Spectrum of Subsolid Nodule Biology. Persistent subsolid nodules almost always represent indolent forms of lung adenocarcinoma, progressing slowly from atypical adenomatous hyperplasia through in-situ and minimally invasive forms. Their slow-growing nature makes them the primary source of potential overdiagnosis and overtreatment in lung cancer screening programs.
Defining Overdiagnosis. Overdiagnosis refers to screen detection and treatment of cancer that would not have affected the patient's life expectancy -- either because the tumor is too indolent to cause harm or because competing comorbidities mean the patient will die from another cause before the cancer becomes clinically significant. For SSN, overtreatment is a real and significant harm to avoid.
Evidence That Surveillance Is Safe. Long-term active surveillance of SSN has proven to be a safe strategy for avoiding overtreatment. In a MILD trial study following 400 SSN over a median of 9 years, SSN never caused death -- while clinically relevant lung cancers mainly developed at other locations. The bioMILD trial subsequently confirmed these findings in a clinical setting.
When SSN Do Progress. The criterion for reclassifying an SSN as positive is demonstrable growth of the solid component -- either newly developing or increasing in size -- or the development of suspicious morphological changes such as bubble-like lucencies. Complex cysts including multilocular cysts and lung cancers associated with cystic airspaces represent a distinct category that requires specific identification, as these often represent invasive forms of lung cancer.
Managing Incidental Findings. Lung cancer screening CT scans frequently reveal findings unrelated to lung cancer, including pulmonary, cardiovascular, and gastrointestinal abnormalities. The reported prevalence of incidental findings varies from 1% to 94% across studies -- reflecting the absence of standardized reporting procedures. Most incidental findings are benign and have little clinical significance.
What Should and Should Not Be Reported. LDCT is not a reliable technique for evaluating solid organs such as the liver, kidney, or breast, which are only partially visualized. Systematic assessment of these organs is not recommended. However, severe coronary artery calcifications -- identified using the Shemesh score -- should be reported given the shared smoking risk factor for both lung cancer and cardiovascular disease. Very suspicious findings warrant reporting based on radiologist judgment.
The Cost-Effectiveness Imperative. Limiting the description of additional findings is essential if LCS programs are to remain cost-effective. Reporting every incidental finding triggers additional diagnostic workups, procedures, and patient anxiety -- generating costs that may erode the population-level benefit of the screening program itself. Evidence-based criteria for selective reporting are necessary.
Structured Reporting Templates. ESTI has developed a comprehensive but simplified structured report template for LDCT screening that includes demographics, acquisition technical details, nodule characteristics, morphology classification, and links to nodule malignancy risk calculation tools. Structured reporting promotes consistency, facilitates program auditing, and optimizes quality assurance across screening centers.
The Complete Picture. A successful lung cancer screening program requires alignment of multiple elements: appropriate patient selection using age and smoking history criteria, technically compliant low-dose CT acquisition, expert radiologist reading with AI support, standardized nodule management protocols, and disciplined structured reporting. Failure in any one element can compromise the program's safety or effectiveness.
The Role of AI Going Forward. AI is positioned to play an increasingly important role in LCS. AI can assist with automated nodule detection, volumetric measurement, malignancy risk estimation, and data mining for program quality monitoring. As AI tools improve and validation evidence accumulates, their role is expected to expand from radiologist aid to a more integrated component of the screening workflow.
Radiologist Leadership. Radiologists must take an active role in guiding and optimizing LCS practices. Qualification through appropriate training programs such as the ESTI LCS Certification Project is essential to ensuring reading quality. Program performance should be continuously monitored through structured auditing processes to identify and address deviations from standard protocol.
Balancing Benefit and Harm. The ultimate goal of lung cancer screening is to reduce lung cancer mortality while minimizing harms -- including radiation exposure, unnecessary invasive procedures for benign nodules, patient anxiety, and overdiagnosis. The ESTI recommendations balance sensitivity for cancer detection against specificity and clinical utility, providing a framework for responsible, evidence-based screening implementation.