• aIcahn School of Medicine at Mount Sinai bDepartment of Radiology cInstitute of Transitional Epidemiology dDepartment of Medicine, Weill Cornell Medical College, New York, New York, USA.

The aim of this study was to assess the impact of the regimen of screening on the frequency of early diagnosis and resection in two computed tomography screening programs. The stage and size distribution of all screen-diagnosed lung cancers was compared. A total of 775 patients in the International Early Lung Cancer Action Program (I-ELCAP) and 664 patients in the National Lung Screening Trial (NLST) were screen-diagnosed; that is, resulting from a positive result requiring further diagnostic workup. The frequency of stage I diagnoses, resections, tumor size, and lung cancer-specific survival were determined. Cox regression was used to identify the key determinants of lung cancer cure. The frequency of clinical stage I lung cancer in I-ELCAP was 82%, and in the NLST it was 67% (P<0.0001). The frequency of stage I (pathologic and clinical if not resected) was 78% in I-ELCAP and 62% in the NLST (P<0.0001). Surgical resection was performed in 86% (664/755) in I-ELCAP and 76% (492/644, P<0.0001) in the NLST. The average tumor size was 17 mm in I-ELCAP and 23 mm in the NLST (P<0.0001). The 5-year survival rate was 83% in I-ELCAP and 62% in the NLST (P<0.0001). Cox regression showed that I-ELCAP provided a 50% better survival benefit than the NLST and that stage I and resection were key determinants of survival, independent of age, smoking history, and tumor size. The higher frequency of stage I disease and resection and smaller tumor size resulted in a significantly higher survival rate in I-ELCAP than in the NLST. These differences strongly support the importance of a specified regimen of screening, as alternative explanations have been addressed.