Lung Cancer Screening Using Low-Dose Spiral CT

Abstract & Commentary

Synopsis: In this nonrandomized screening study of 7956 individuals, the study found that low-dose spiral CT was a good screening test for detecting early lung cancer. This study also showed that annual screening should be done only in selected individuals.

Source: Nawa T, et al. Chest. 2002;122:15-20.

Lung cancer is currently the leading cause of death due to malignancy in the United States, but screening has been a major issue. In the past, the opinion was that chest roentgenograms and sputum cytology could prove to be useful for early detection of lung cancer, but several studies failed to demonstrate efficacy and CXR screening subsequently has fallen out of favor.

This was a prospective, nonrandomized, institutional study in Japan screening 7956 individuals for primary lung cancer, which extended for 29 months. Based on the results of this thoracic CT screening, the following data were extracted: a) prevalence of lung cancer detected at baseline screening and its incidence at the repeat screening; b) detectability of small, early lung cancer; and c) selection of suitable candidates. Demographics were as follows: age range 50-69, 77.1% of males in the study population were smokers, 4.3% of females in the study population were smokers, 79% of the total population were male and 21% female. Nodules that showed extensive calcification or clustered in the same lobe in groups of 6 or greater were considered to be benign and excluded from the study.

Noncalcified solid pulmonary nodules (SPN) measuring 8-10 mm were examined with detailed CT (thin cut CT chest) 1 month later, and if the size of the nodule was 5-7 mm, an annual repeat CT screening was recommended. If the detailed CT demonstrated a noncalcified SPN 8-10 mm in size, a detailed CT was repeated every 3 months for 6 months, and if no growth was noted, annual screening CT was recommended. If at any point growth of the nodule was noted, a further study was recommended: bronchoscopy, video-assisted thoracoscopy (VATS) biopsy, and/or serum tumor markers.

Results of the baseline screening showed that 2099 of 7956 (26.3%) patients had noncalcified SPN, and 541 of 2099 patients (6.8% of 7956) had a detailed CT. Subsequent bronchoscopy was done in 64 of 541 (11.8%) individuals and VATS biopsy was subsequently performed in 51 of these 64 patients (9.4% of 541 individuals with detailed CT). Of the 541 (6.7%) individuals who were identified to have noncalcified nodules measuring 8 mm or greater on a detailed CT, 36 (6.2%) were diagnosed with primary lung cancer. Therefore, the prevalence of primary lung cancer at baseline was 0.44% for the 7956 individuals. Of those patients diagnosed with lung cancer, 9 (25%) individuals were smokers.

The staging of lung cancer at baseline screening ranged as follows: stage IA (28/36; 77.7%); stage IB (3/36; 8.3%); stage IIA (3/36; 8.3%); stage IIB (1/36; 2.7%); and stage IIIA (1/36; 2.7%). Histologically, the tumors were determined to be well-differentiated adenocarcinoma (27/36), moderately differentiated adenocarcinoma (7/36), large cell carcinoma (1/36), and carcinoid (1/36).

A follow-up screening with low-dose CT was repeated a year later in 5568 individuals who had no evidence of noncalcified pulmonary nodules at baseline, and 148 of 5568 individuals were recommended to have a detailed CT for noncalcified nodules greater than or equal to 8 mm. Of those individuals with detailed CT, 7 of 148 (4.7%) had additional diagnostic tests (bronchoscopy and/or VATS), and 4 of 148 (2.7%) were diagnosed to have primary lung cancer. The incidence of primary lung cancer was 0.07%. All of these individuals were smokers.

Primary lung cancer detected on the annual screening follow-up was stage IA in all 4 cases, with 2 well differentiated and 2 moderately differentiated adenocarcinomas.

Comment by David Ost, MD, & Andreas Kyprianou, MD

Previous authors have found no improvement in mortality of lung cancer by screening patients with chest roentgenograms and sputum cytology,1 but recently encouraging results of low-dose spiral CT have been reported.2 Nawa and colleagues have demonstrated that low-dose spiral CT can be a good screening test for early primary lung cancer, especially for adenocarcinoma and large cell cancer, but not for squamous cell or small cell lung cancer. Further prospective studies are warranted to demonstrate improved mortality and to assess cost effectiveness of using low-dose spiral CT as a screening test for lung cancer. This study supports the theory that annual lung cancer screening may be useful when applied to selected populations. In particular, the limited data available suggest that it might be targeted at individuals with a history of smoking.

Dr. Ost, Assistant Professor of Medicine, NYU School of Medicine, Director of Interventional Pulmonology, Division of Pulmonary and Critical Care Medicine, Northshore University Hospital, Manhasset, NY, is Associate Editor of Internal Medicine Alert. Dr. Kyprianou is a Fellow in Pulmonary/Critical Care Medicine, Northshore University Hospital, Manhasset, NY.


1. Flehinger BJ, et al. Am Rev Respir Dis. 1984;130:555-560.

2. Kaneko M, et al. Radiology. 1996;201:798-802.