Virtual Library

Start Your Search

Ioannis Tsmardinos

Author of

  • +

    MA20 - Implementation of Lung Cancer Screening (ID 923)

    • Event: WCLC 2018
    • Type: Mini Oral Abstract Session
    • Track: Screening and Early Detection
    • Presentations: 1
    • Moderators:
    • Coordinates: 9/25/2018, 15:15 - 16:45, Room 206 F
    • +

      MA20.02 - “Reduced” HUNT Lung Cancer Model for Predicting Lung Cancer in the Prospective Danish Lung Cancer Screening Study - Comparison with the NLST. (ID 12802)

      15:20 - 15:25  |  Author(s): Ioannis Tsmardinos

      • Abstract
      • Presentation
      • Slides


      Risk prediction is important for selection of individuals for lung cancer screening programs. We have recently published a validated calculator, the HUNT Lung Cancer Risk Model ( for all ages and smoking patterns.

      The Danish Lung Cancer Screening Trial (DLCST) was a randomized prospective study that included 4104 heavy smokers with a median age of 57.6, 33.8 pack-years and maximum 9 years quit time and 10 years follow-up. We tested the value of the HUNT model in this prospective Danish lung cancer study and compared with the NLST.

      a9ded1e5ce5d75814730bb4caaf49419 Method

      The DLCST study only had 5 of the 7 variables in the original HUNT Model, so we trained a “reduced” HUNT Model in the Norwegian HUNT2 cohort of 12 091 ever-smokers ages between 49-71 years (as the age span in the Danish cohort) based on age, pack-years, smoking intensity, quit time and BMI where sex was added for adjustment. The model was applied blindly in the 4051 individuals of the Danish cohort that had all 5 variables.

      4c3880bb027f159e801041b1021e88e8 Result

      In the population selected by the "reduced" HUNT Model, 148/149 (99.33%) lung cancer cases were predicted (sensitivity 99.33%, negative predictive value 99.23%), thus only one individual that developed lung cancer was lost among those 52 not selected for screening. If the the NLST criteria were used (age 55-74, >30 pack-years and <15 years quit time), less than half of the Danish cohort (n=1870 (46.2%)) would have been considered eligible for screening, and 104/149 (69.80%) lung cancer cases would have been predicted. With these criteria, one would lose 45 (32.7%) lung cancer cases, and the sensitivity would be lower (69.80%).

      8eea62084ca7e541d918e823422bd82e Conclusion

      We were able to predict 99.33% of those that were diagnosed with lung cancer in 10 years, only one future lung cancer case was not included. Therefore, even the “reduced” HUNT model was highly sensitive in selecting persons at high risk for lung cancer in a screening cohort. The objection one could have for preferring the NLST criteria is that one screened about half of this population but at the same time lost 1/3 of the future lung cancers. In a health system that can afford to screen more people than those included by the NLST criteria, the HUNT model may be useful, preferably the validated HUNT Lung Cancer Model, for the selection of high risk individuals to a screening programme.


      Only Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login, select "Add to Cart" and proceed to checkout. If you would like to become a member of IASLC, please click here.

      Only Active Members that have purchased this event or have registered via an access code will be able to view this content. To view this presentation, please login or select "Add to Cart" and proceed to checkout.