As we prepare to receive the details of the National Lung Screening Trial (NLST) results over the coming months, it is important to remind ourselves that CT screening of the lungs provides secondary prevention for lung cancer and other tobacco associated diseases. Smoking cessation will continue to provide primary prevention for these life threatening diseases. Patients who received CT screening for lung cancer at Brigham and Women's Hospital prior to NLST more often than not used negative results to justify continuing their smoking behavior. Thus developed a chasm between primary and secondary prevention.
Over the past decade, through lung screening trials and the COPDGene™ cohort study, radiologists, pulmonologists and physiologists have increased the vocabulary for CT screening of lung parenchyma. Qualitative and quantitative measures must inform one another for descriptions of subtle abnormalities to enter into clinical practice. Efforts to date have largely focused on emphysema, larger airways, and airway obstruction. Early pilot investigations of more subtle differences in lung parenchyma indicate that we have not reached the limit of diagnostic information available from CT scan images.
CT scan technology continues to evolve resulting in a variety of differences in images over time. Technical factors that affect patient dose of ionizing radiation are responsible for a portion of the differences. Some differences are based on imaging system physics although departures from Hounsfield standards for air and water can be difficult to explain. This limits the direct comparability of quantitative measurements over time without even considering unavoidable differences related to lung expansion and slice selection. Radiologists become accustomed to the images that they see most frequently and can provide valid subjective comparisons across images that appear quite different from one another. The differences required for detection increase with increasing disparity between images. This and the lack of detailed clinical history for correlation have limited clinical evaluation of subtle features and changes, even over short periods of time, that may be related to continued smoking. Dr. Shaker and his colleagues have provided systematic insight into the effect of smoking cessation on the CT appearance of lung parenchyma. While the pathophysiologic correlation is limited, the difference could provide motivation and reinforcement of smoking cessation in clinical practice. As it is unlikely to represent progression of emphysema, the findings reported will be important to disseminate to radiologists as CT lung screening enters mainstream clinical practice. The possibility of relatively rapid improvement in lung function following smoking cessation that may be represented by the decrease in lung attenuation would also be a welcome additional source of motivation for smoking cessation. The ability to utilize the CT scan to help support the individual who is struggling to quick smoking represents a turning point in the relationship between primary and secondary prevention of lung cancer. For lung screening to succeed in the long-term, it will be imperative to develop the relationships between smoking, smoking cessation and the appearance of lung parenchyma on CT. The reinforcement of primary prevention through secondary prevention is furthered by Shaker et al in this issue.
Human nature and the technological evolution of CT scanning equipment must both be brought into the discussion rather than swept under the rug. The detection of subtle changes over time, due to a variety of factors including differences in respiratory cycle, continued smoking, and smoking cessation poses challenges to both subjective observation by subspecialty-trained radiologists and the objective quantitative measurement of lung density and airway thickness. Further research should be undertaken to understand the subtle differences in CT appearance of lung parenchyma. Analytic metrics need to be developed that will remain applicable over time and across scanners.