Variations in lung cancer risk among smokers

Our plain-English summary of Variations in lung cancer risk among smokers, by P.B. Bach, M.W. Kattan, M.D. Thornquist, et al (Journal of the National Cancer Institute, 2003).

The study

• Authors: P.B. Bach, M.W. Kattan, M.D. Thornquist, M.G. Kris, R.C. Tate, M.J. Barnett, L.J. Hsieh, C.B. Begg
• Journal: Journal of the National Cancer Institute
• Published: 2003;95(6):470-8
• Cited by: 491 articles, as counted by NIH iCite
• Read it: PubMed · Publisher

Abstract

Background: Although there is no proven benefit associated with screening for lung cancer, screening programs are attracting many individuals who perceive themselves to be at high risk due to smoking. We sought to determine whether the risk of lung cancer varies predictably among smokers.

Methods: We used data on 18 172 subjects enrolled in the Carotene and Retinol Efficacy Trial (CARET)-a large, randomized trial of lung cancer prevention-to derive a lung cancer risk prediction model. Model inputs included the subject's age, sex, asbestos exposure history, and smoking history. We assessed the model's calibration by comparing predicted and observed rates of lung cancer across risk deciles and validated it by assessing the extent to which a model estimated on data from five CARET study sites could predict events in the sixth study site. We then applied the model to evaluate the risk of lung cancer among smokers enrolled in a study of lung cancer screening with computed tomography (CT).

Results: The model was internally valid and well calibrated. Ten-year lung cancer risk varied greatly among participants in the CT study, from 15% for a 68-year-old man who has smoked two packs per day for 50 years and continues to smoke, to 0.8% for a 51-year-old woman who smoked one pack per day for 28 years before quitting 9 years earlier. Even among the subset of CT study participants who would be eligible for a clinical trial of cancer prevention, risk varied greatly.

Conclusions: The risk of lung cancer varies widely among smokers. Accurate risk prediction may help individuals who are contemplating voluntary screening to balance the potential benefits and risks. Risk prediction may also be useful for researchers designing clinical trials of lung cancer prevention.

In plain English

This study built and tested a model to predict a smoker's chance of being diagnosed with lung cancer within 10 years. The authors used data on 18 172 current or former smokers in CARET, a US lung cancer prevention trial. Of these, 3918 were asbestos-exposed men, mainly shipyard or construction workers. The model used age, sex, asbestos exposure history and smoking history. A history of asbestos exposure, meaning at least 5 years in a high-risk trade with first exposure 15 or more years earlier, was associated with higher lung cancer risk (hazard ratio 1.24, 95% CI 1.04 to 1.48). The type of asbestos was not considered. Predictions closely matched observed cases at held-out study sites. In a Mayo Clinic CT screening study, one example at the 5th percentile of risk was a 51-year-old woman who quit 9 years earlier: her predicted 10-year risk was 0.80%, assuming she stayed abstinent. An example at the 95th percentile, a 68-year-old man smoking two packs a day, had a predicted 15% if he kept smoking. The authors concluded that lung cancer risk varies widely among smokers. They noted the model needs further validation.

We wrote this plain-English summary from the full article with the help of AI, and a second AI model checked each sentence against the article. It is not the authors' wording; the abstract and the article are their own account.

What the study did and found

• Design: Development and validation of a lung cancer risk prediction model (Cox proportional hazards regression) using data from a randomized prevention trial, then applied to participants in a CT screening study
• Who was studied: 18 172 current or former smokers enrolled in CARET (14 254 heavy smokers, men and women aged 50–69, and 3918 asbestos-exposed men) at six US sites, enrolled 1985–1994 and followed to February 25, 2002. The model was then applied to 300 randomly selected participants in a Mayo Clinic (Rochester, MN) low-dose CT screening study, of whom 201 met National Lung Screening Trial entry criteria.
• Exposure: Smoking history (duration, cigarettes per day, years since quitting) and a history of occupational asbestos exposure, defined as first occupational exposure 15 or more years earlier and at least 5 years in a high-risk trade (e.g., asbestos worker, insulator, shipyard trades, plumber/pipefitter). Type of asbestos was not considered.
• Main results:
  • In the CARET cohort, a history of asbestos exposure was associated with an independent increase in lung cancer risk (HR 1.24, 95% CI 1.04 to 1.48)
  • The model's cross-validated concordance index was 0.72
  • When the model was derived from five CARET sites, observed lung cancer rates at the held-out sixth site closely matched predicted rates
  • An example participant at the 5th percentile of risk in the Mayo Clinic study, a 51-year-old woman who smoked one pack per day for 28 years and quit 9 years earlier, had a predicted 10-year lung cancer risk of 0.80%, assuming she remains abstinent
  • An example participant at the 95th percentile of risk in the Mayo Clinic study, a 68-year-old male current smoker (two packs per day for 50 years), had a predicted 10-year risk of 11% if he quit immediately and 15% if he continued smoking
  • Among the 201 Mayo Clinic participants meeting NLST entry criteria, the lowest-risk quarter was expected to account for about 8% of lung cancer cases and the highest-risk quarter for about 50%
  • The model showed no convincing association between sex and lung cancer risk (HR 0.94, 95% CI 0.92 to 1.08)
  • The authors did not observe an additional independent benefit from longer duration of quitting, beyond differences in smoking duration
• The authors' conclusion: The authors concluded that lung cancer risk varies widely among smokers, and that accurate risk prediction may help people considering screening and researchers designing prevention trials
• Limitations the authors state:
  • The model does not distinguish between histologic types of lung cancer
  • The model applies only to people aged 50 or older with a smoking history
  • The model needs further validation
  • Subjects were clinical trial participants and are not perfectly representative of the general population
  • Systematic differences by sex in how other exposures were recorded could have masked a small effect of sex
  • Type of asbestos exposure was not considered as a predictor because it was not easily assessable or not recorded in CARET

Topics

The National Library of Medicine indexes this study under subjects our site covers: Lung cancer.

Medical terms in this study, explained in our Lexicon: Lung Cancer, Asbestos Exposure, Low-Dose CT Screening, Clinical Trial.

Related studies

• Yield of Low-Dose Computerized Tomography Screening for Lung Cancer in High-Risk Workers: The Case of 7189 US Nuclear Weapons Workers — S.B. Markowitz, A. Manowitz, et al. (2018)
• External validation and recalibration of the Brock model to predict probability of cancer in pulmonary nodules using NLST data — A. Winter, D.R. Aberle, et al. (2019)
• Methylation-derived Neutrophil-to-Lymphocyte Ratio and Lung Cancer Risk in Heavy Smokers — L. Grieshober, S. Graw, et al. (2018)
• Estimated intake of vitamin D and its interaction with vitamin A on lung cancer risk among smokers — T.Y. Cheng, G.E. Goodman, et al. (2014)
• Current status of screening for malignant pleural mesothelioma — H.I. Pass, M. Carbone (2009)
• Factors determining early adherence to a lung cancer screening protocol — U. Montes, L.M. Seijo, et al. (2007)

Browse more research in Journal of the National Cancer Institute, or return to medical research.