NLST Analysis: Lung Screening–Detected Abnormalities Other Than Cancer Result in Smoking Cessation
In a study reported in the Journal of the National Cancer Institute, Tammemägi et al assessed smoking cessation rates among participants undergoing chest x-ray or computed tomography (CT) screening for lung cancer in the National Lung Screening Trial (NLST). Among patients without a subsequent diagnosis of lung cancer, they found that screening abnormalities were significantly associated with smoking cessation. This finding suggests that lung cancer screening programs may provide opportunities to reduce smoking rates.
Study Details
In the study, data from participants in the NLST (2002–2009) were used in multivariable longitudinal regression models to predict annual smoking cessation in 15,489 subjects who were current smokers at study entry and did not receive a diagnosis of lung cancer during the study. Screening was conducted at years 0, 1, and 2, with annual follow-up at years 3 to 7. The NSLT showed that annual low-dose CT screening reduced lung cancer mortality by 20% compared with chest x-ray screening
Overall, at study year 3, 23.5% of subjects were no longer smoking. The highest proportions of subjects who remained smokers at ≥ 1-year intervals after a screening were those with normal screening results, followed in order by those with a screen showing a minor abnormality not suspicious for lung cancer, those with a major abnormality not suspicious for lung cancer and those with a screen that was suspicious for lung cancer but was unchanged from the previous screen, and those with a screen suspicious for lung cancer that represented a new or changing finding.
Year-to-Year Impact
Among subjects with a normal finding at year 0 screening, 87.4% remained smokers at year 1 screening. By comparison, rates were significantly lower among patients with major abnormality not suspicious for cancer (85.0% still smokers, odds ratio [OR] = 0.81) and among those with an abnormality suspicious for lung cancer (81.7% still smokers, OR = 0.64).
Among subjects with a normal finding at year 1 screening, 83.1% remained smokers at year 2 screening; by comparison, smoking rates were significantly lower in those with findings suspicious for lung cancer but stable from year 1 screen (79.2%, OR = 0.78) and those with a new finding suspicious for lung cancer (74.8%, OR = 0.61).
Among subjects with a normal finding at year 2 screening, 78.3% remained smokers at year 3; by comparison, significant reductions in smoking rates were found among subjects with an unsuspicious minor abnormality at year 2 (76.3%, OR = 0.89), those with an unsuspicious major abnormality (73.3%, OR = 0.76), those with a stable suspicious finding (73.9%, OR = 0.79), and those with a new/changed suspicious finding (71.9%, OR = 0.71).
Effect of Longer Durations
Among subjects with normal findings on year 2 screening, 73.6% remained smokers at year 4 follow-up. By comparison, there were significant reductions in smoking rates at year 4 among those with an unsuspicious minor abnormality (71.8%, OR = 0.91), those with an unsuspicious major abnormality (68.4%, OR = 0.77), those with a stable suspicious finding (69.9%, OR = 0.83), and those with a new/unstable suspicious finding (68.0%, OR = 0.76).
At years 5, 6, and 7, the smoking rates among subjects with normal results at year 2 screening decreased from 68.1% to 61.8%; by comparison, rates nonsignificantly decreased from 67.4% to 60.7% in subjects with an unsuspicious minor abnormality, from 64.5% to 60.9% in those with an unsuspicious major abnormality, and from 65.4% to 58.1% in those with a stable suspicious finding. Among those with a new/unstable suspicious finding at year 2 screening, decreases in smoking rates compared with those with normal findings were significant at year 5 (63.6%, OR = 0.82), year 6 (58.1%, OR = 0.76), and year 7 (56.7%, OR = 0.81).
Screening Abnormalities Significantly Predict Cessation
A final multivariable logistic model included adjustment for sociodemographic factors (age, sex, race/ethnicity, education as an indicator of socioeconomic circumstance, and marital status), exposures (alcohol consumption, cigarette, cigar and pipe smoking histories, and secondhand smoke exposures), and medical history (body mass index [BMI], family history of lung cancer, personal history of cancer, history of comorbidities), as well as for study year, study center, and randomization group (x-ray or CT screening).
On this model, compared with normal screening findings, findings of an unsuspicious minor abnormality (OR = 0.91, P = .005), an unsuspicious major abnormality (OR = 0.81, P < .001), a stable suspicious abnormality (OR = 0.78, P < .001), or a new/unstable suspicious finding (OR = 0.66, P < .001) were independently predictive of smoking cessation. Other significant predictors for not smoking were increasing age, mixed race (vs white), increased education level, married status, higher BMI, lower smoking intensity and duration, absence of secondhand smoke at home, and past or present pipe and cigar smoking.
Outcomes by X-Ray or CT Screening
Analysis according to randomization group showed that compared with subjects in the x-ray group with normal findings, there was a significant reduction in smoking among those with an unsuspicious minor abnormality (adjusted OR = 0.86, P < .001), a borderline significant reduction in those with an unsuspicious major abnormality (OR = 0.81, P = .06), and significant reductions in those with a stable suspicious finding (OR = 0.72, P = .02) and those with an new/unstable suspicious finding (OR = 0.71, P < .001).
Compared with subjects in the CT group with a normal finding, there was a nonsignificant increase in smoking (adjusted OR = 1.01, P = .88) in those with an unsuspicious minor abnormality, a borderline significant decrease in those with an unsuspicious major abnormality (OR = 0.87, P = .06), and significant reductions in those with a stable suspicious finding (OR = 0.84, P = .01) or a new/unstable suspicious finding (OR = 0.69, P <.001).
The investigators concluded, “Smoking cessation is statistically significantly associated with screen-detected abnormality. Integration of effective smoking cessation programs within screening programs should lead to further reduction in smoking-related morbidity and mortality.”
Martin C. Tammemägi, PhD, of Brock University, Ontario, is the corresponding author for the Journal of the National Cancer Institute article.
The National Cancer Institute is the funding source for the NLST. The study authors reported no potential conflicts of interest.
The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.
