BCCRI-LunCan (BCCRI) British Columbia Cancer Research Institute

Overview

The BCCRI-LungCan Model is a discrete-state microsimulation model. The purpose of the model was to evaluate the effect of screening on lung cancer incidence and mortality, survival outcomes, screening-related harms, and quality of life. In collaboration with the other CISNET Lung groups, it was used for the 2021 USPSTF lung cancer screening decision analyses,¹ the 2023 lung cancer screening guidelines update from the American Cancer Society,² the effect of risk-based screening strategies, and to assess the cost-effectiveness of various LDCT screening strategies in the United States.^3,4 Furthermore, this model has been extended to incorporate a smoking cessation component to evaluate the health outcomes and cost-effectiveness of smoking cessation interventions in the context of lung cancer screening.^5-7 Additionally, the model was used to evaluate the benefits and harms of lung cancer screening strategies with adaptive schedules selected based on individual risks and life expectancy.⁸

The BCCRI-LungCan Model is a combination of a multistage carcinogenesis model and a discrete-state microsimulation model. This model consists of three main components: population, natural history, and screening. A sub-component under the screening component—smoking cessation intervention component—was developed to evaluate smoking cessation interventions within lung screening program. The population component utilized the Smoking History Generator to simulate input population for the natural history and screening components.

The natural history component simulates individual lung cancer–oriented life events in the absence of screening given the individual’s smoking history. The age at clinical detection of lung cancer (dose-response) is simulated through a Two-Stage Clonal Expansion (TSCE) model.⁹ The TSCE model assumes that two mutation (rate-limiting) events are required for the initiation of premalignant lung tumors, and it explicitly models the dynamics of premalignant and malignant tumors. The TSCE model was fitted to lung cancer incidence in the Nurses' Health Study and the Health Professionals Follow-up Study using a likelihood-based approach. If an individual develops clinically diagnosed lung cancer, the natural history model also simulates the age at lung cancer onset, histology, stage at diagnosis, preclinical sojourn time for each stage, and age at lung cancer death. Alternative versions of this model use different dose-response models (Bach, PLCOm2012, and LCRAT models). Lung cancer histology is classified into four main groups: small cell, adenocarcinoma, squamous, and other. Histology is simulated using a multinomial logistic regression accounting for sex, age, and smoking exposure model based on the PLCO. Preclinical sojourn times for each stage follow a Weibull distribution with shape and scale parameters depending on sex, stage, and histology. The preclinical sojourn times are simulated for all stages for each clinically diagnosed lung cancer case and used in the screening component to model the effect of screening. Lung cancer–specific survival time conditioned on sex, age at diagnosis, histology, and stage is estimated by using cure models with lognormal survival distributions, which were fitted to lung cancer survival data in SEER 18.

The outputs from the natural history component serve as inputs for the screening component to simulate the effect of screening on lung cancer incidence, stage, and survival. The model uses sensitivity and specificity rates to simulate screening results: true positive, false positive, or negative. Sensitivity varies by screening round, lung cancer histology, and stage. These sensitivities, originally based on NLST, were adjusted to conform to Lung-RADS criteria by multiplying them by a scaling factor, given by the ratio of the overall sensitivity from Lung-RADS over that from NLST. Specificity rates by screening round were also based on a retrospective analysis of NLST outcomes using the Lung-RADS criteria.

The model simulates additional screening outcomes such as the number of follow-up tests and potential complications based on NLST rates. The rates of specificity and sensitivity for screening simulation were chosen based on other published literature. For true-positive cases, lung cancer–specific survival time and stage burden were updated given the stage and age at diagnosis. The model also simulates radiation-induced lung cancer from CT scan exposure.

References

1. Meza R, Jeon J, Toumazis I, Ten Haaf K, Cao P, Bastani M, Han SS, Blom EF, Jonas DE, Feuer EJ, Plevritis SK, de Koning HJ, Kong CY. Evaluation of the Benefits and Harms of Lung Cancer Screening With Low-Dose Computed Tomography: Modeling Study for the US Preventive Services Task Force. JAMA. 2021;325(10):988-97.
2. Meza R, Cao P, de Nijs K, Jeon J, Smith RA, Ten Haaf K, de Koning H. Assessing the impact of increasing lung screening eligibility by relaxing the maximum years-since-quit threshold: A simulation modeling study. Cancer. 2024;130(2):244-55.
3. Toumazis I, de Nijs K, Cao P, Bastani M, Munshi V, Ten Haaf K, Jeon J, Gazelle GS, Feuer EJ, de Koning HJ, Meza R, Kong CY, Han SS, Plevritis SK. Cost-effectiveness Evaluation of the 2021 US Preventive Services Task Force Recommendation for Lung Cancer Screening. JAMA oncology. 2021;7(12):1833-42. doi: 10.1001/jamaoncol.2021.4942. PubMed PMID: 34673885; PMCID: PMC8532037.
4. Ten Haaf K, Bastani M, Cao P, Jeon J, Toumazis I, Han SS, Plevritis SK, Blom EF, Kong CY, Tammemagi MC, Feuer EJ, Meza R, de Koning HJ. A Comparative Modeling Analysis of Risk-Based Lung Cancer Screening Strategies. J Natl Cancer Inst. 2020;112(5):466-79. doi: 10.1093/jnci/djz164. PubMed PMID: 31566216; PMCID: PMC7225672.
5. Cao P, Smith L, Mandelblatt JS, Jeon J, Taylor KL, Zhao A, Levy DT, Williams RM, Meza R, Jayasekera J. Cost-Effectiveness of a Telephone-Based Smoking Cessation Randomized Trial in the Lung Cancer Screening Setting. JNCI cancer spectrum. 2022;6(4).
6. Meza R, Cao P, Jeon J, Taylor KL, Mandelblatt JS, Feuer EJ, Lowy DR. Impact of Joint Lung Cancer Screening and Cessation Interventions Under the New Recommendations of the U.S. Preventive Services Task Force. J Thorac Oncol. 2022;17(1):160-6.
7. Cao P, Jeon J, Levy DT, Jayasekera JC, Cadham CJ, Mandelblatt JS, Taylor KL, Meza R. Potential Impact of Cessation Interventions at the Point of Lung Cancer Screening on Lung Cancer and Overall Mortality in the United States. J Thorac Oncol. 2020;15(7):1160-9.
8. Cao P, Jeon J, Meza R. Evaluation of benefits and harms of adaptive screening schedules for lung cancer: A microsimulation study. J Med Screen. 2022;29(4):260-7.
9. Meza R, Hazelton WD, Colditz GA, Moolgavkar SH. Analysis of lung cancer incidence in the Nurses' Health and the Health Professionals' Follow-Up Studies using a multistage carcinogenesis model. Cancer Causes Control. 2008;19(3):317-28.