University of Michigan Lung Cancer Smoking Model
The University of Michigan Lung Cancer Smoking Model (UM-LCSm) was formerly known as the FH-LC Model (Fred Hutchinson Cancer Research Center). The UM-LCSm is an effective tool for evaluating lung cancer trends in the US population and the effects of possible interventions (1). The model was developed in two steps. First, a natural history model that describes lung cancer incidence or mortality as functions of smoking history and age was derived, estimating the model parameters by fitting it to prospective cohort data of smoking histories and lung cancer incidence/mortality (2-4). Second, the natural history model was embedded in an Age-Period-Cohort (APC) population model (5), and relative lung cancer mortality rates by calendar-year and birth-cohort (period and cohort effects, respectively) were estimated by calibration against US lung cancer mortality. The UM-LCSm has been used to estimate the number of lung cancer deaths prevented in the US due to tobacco control efforts implemented since the first Surgeon General’s Report (SGR) on smoking and health was issued in 1964 (1,4).
The UM-LCSm model is based on the biologically based two-stage clonal expansion (TSCE) model that relates individual smoking histories to the age-specific risk of lung cancer incidence or mortality (6). The TSCE model is a stochastic model that represents the process of carcinogenesis in three phases. In the first phase (initiation), a susceptible stem cell acquires one or more mutations resulting in an initiated cell, which has partially escaped growth control. In the second phase (promotion), initiated cells undergo clonal expansion, either spontaneously or in response to endogenous or exogenous promoters. Finally, in the third phase (malignant conversion), one of the initiated cells acquires further mutational changes leading to a malignant cell.
To model the effects of smoking on lung cancer risk, the model initiation, promotion, and malignant transformation parameters are assumed to be altered during periods of smoking exposure through parametric dose-response relationships. This dose-response relationship links the individual smoking history to the cell kinetic parameters in the TSCE model.
Model calibration consists of estimating dose-response parameters that best represent the effects of individual smoking histories in relation to lung cancer initiation, promotion, malignant conversion, and in turn, incidence or mortality. Using likelihood-based methods, the TSCE model was calibrated to lung cancer data in four US smoking cohorts: lung cancer mortality in the American Cancer Society Cancer Prevention Studies I and II (CPS-I and CPS-II)(2), and lung cancer incidence and mortality in the Nurses’ Health Study (NHS) and Health Professionals’ Follow-up Study (HPFS) (3,4).
As mentioned above, the calibrated TSCE models were then embedded into an APC model replacing the non-parametric age effects of the traditional APC models by the TSCE model hazard (i.e., model age-specific incidence). Period and cohort effects were then estimated by calibrating the model predictions to lung cancer mortality in the US population, using microsimulation of individual smoking histories in the US from CISNET’s Smoking History Generator (SHG) (7).
Tip: Hover your cursor over the dashed attribute links below for more information. View the details of this model in a grid with other lung models.