Proven Efficacy, Yet a Persistent ‘Implementation Gap’
Lung cancer screening using Low-Dose Computed Tomography (LDCT) is no longer a novel concept. Its efficacy has been well established through large-scale randomized controlled trials (RCTs).
The U.S. National Lung Screening Trial (NLST) demonstrated a 20% reduction in lung cancer mortality, while the European NELSON trial similarly reported a 24% reduction.
In contrast, the global implementation of cancer screening programs remains uneven. Even for cancers such as breast and cervical cancer—both recommended by the World Health Organization (WHO) for population-based screening—national screening programs are not universally in place. Current estimates suggest that organized cervical cancer screening programs exist in approximately 69% of countries, and breast cancer screening programs in around 63%.
Against this backdrop, lung cancer screening faces an even greater implementation challenge.
While the clinical effectiveness of LDCT screening is well established, its adoption at a national or regional level remains limited in many healthcare systems.
The ZORALCS study, conducted in the Flemish region of Belgium, begins precisely at this point.
Rather than revisiting the question of efficacy, the study asks a more practical one:
Can lung cancer screening be operated sustainably within existing healthcare infrastructures?
The Core Question: "Can the System Handle Reality?"
The goal of the ZORALCS study is not merely to reaffirm the efficacy of screening. It is an 'Implementation Study' designed to verify whether the system functions correctly when the screening program is transplanted from a controlled clinical trial environment into a real-world community healthcare setting.
In Belgium, lung cancer is the second most common cancer and a critical disease causing approximately 5,700 deaths annually. The reason for the absence of a national screening program despite this severity is clear: the inability to solve the challenges of 'Standardization' and 'Reproducibility' of readings when targeting a large-scale population.
AI: Not Just a Tool, But ‘Infrastructure’ for the System
A prerequisite for the success of national screening is 'consistent judgment' that does not rely on the proficiency or condition of individual physicians. To achieve this, the ZORALCS research team adhered to the strict guidelines of the European Society of Thoracic Imaging (ESTI) while establishing a mandatory condition: the introduction of 'State-of-the-art' AI as a Second Reader.
What the researchers demanded from AI was more than simple assistance. It was necessary to convert minute changes, often difficult to confirm with the human eye, into manageable data through the 'quantitative analysis' core to the screening system.
- Precise Nodule Detection and Segmentation: Accurately detects and segments pulmonary nodules in three dimensions, providing a robust technical foundation for downstream quantitative analysis while minimizing the risk of missed findings in lung cancer screening settings.
- Automated Volume Analysis (Volumetry): Provides both conventional 2D diameter measurements and precise 3D volumetric assessments based on accurate nodule segmentation, ensuring continuity with traditional reading practices while enabling reliable longitudinal measurement in accordance with current lung cancer screening guidelines.
- Automated Volume Doubling Time (VDT) Calculation: Automatically calculates nodule growth rates across follow-up scans and supports guideline-based interpretation of nodule growth, applying established clinical frameworks such as ESTI, Lung-RADS, BTS (UK), and K-Lung-RADS. The system further enables flexible customization to align with national lung cancer screening programs and pilot initiatives, including K-NLCSP (Korea), NLCSP (Australia), and German LCS, supporting risk stratification within country-specific screening workflow.
This technology eliminates subjective interpretation in readings and functions as the most realistic and powerful screening infrastructure, ensuring uniform quality even in a multi-center environment.
Three Key Elements for Successful Introduction
The ZORALCS study offers significant implications for institutions considering the introduction of lung cancer screening. The following three elements are essential for successful implementation:
- Standardization: Establishing protocols based on international guidelines (e.g., ESTI).
- Quantification: Securing objective metrics via AI-powered 3D volumetry and VDT analysis.
- Systematization: Integrating AI not as a mere support tool, but as essential infrastructure (Second Reader) for quality control.
Verified Solutions: The Conditions for Feasible Lung Cancer Screening
What Belgium sought to confirm through the ZORALCS study was not just the ‘effect of screening,’ but the ‘conditions’ required to sustain that effect within a real-world medical system.
The criteria set by the research team were rigorous: standardization based on international guidelines, quantitative indicators to minimize reading discrepancies, and a structure capable of maintaining identical quality across multi-center environments. There was a fundamental judgment that without these three prerequisites, lung cancer screening could not be expanded into a national system.
In this context, Coreline Soft’s AI-based analysis technology, utilized in this study, holds significant meaning. It was employed not as a subject to prove technical superiority, but as an ‘implementation enabler’ to ensure the stable operation of the screening system. It served as an infrastructure role, providing standardized reading criteria and quantitative analysis so that screening results would not be swayed by specific institutions or the personal experience of readers.
This approach is not limited to Belgium. In lung cancer screening and pilot projects conducted in leading nations like Germany and Italy, the requirement was not a competition for ‘better algorithms,’ but for a ‘verifiable and reproducible screening structure.’
Ultimately, what is needed for lung cancer screening to be successfully institutionalized is not a performance competition of single technologies.
Only when standardized protocols are combined with an AI infrastructure that consistently supports them can screening truly function as a ‘System.’ The ZORALCS study is a case that has proven this possibility not in theory, but in operational reality.
Reference
- ZORALCS Study: Van der Aalst, C., et al. "Design and rationale of the ZORALCS study: An implementation study of lung cancer screening by low-dose computed tomography coupled to a smoking cessation randomized controlled trial in the Flemish region."
- NLST (20% Mortality Reduction): National Lung Screening Trial Research Team. "Reduced lung-cancer mortality with low-dose computed tomographic screening." New England Journal of Medicine 365.5 (2011): 395-409.
- NELSON (24% Mortality Reduction): de Koning, Harry J., et al. "Reduced lung-cancer mortality with volume CT screening in a randomized trial." New England Journal of Medicine 382.6 (2020): 503-513.
List
