The PSAPC model is a microsimulation model that links a man's prostate-specific antigen (PSA) levels and prostate cancer (PC) progression. PSA levels grow exponentially with rates before and after disease onset (Figure A) estimated using data from the control arm of the Prostate Cancer Prevention Trial1, which screened 9,000 men annually for up to 7 years with an exit biopsy regardless of PSA test results. PSA growth rates are greater for high-grade disease. Risks of onset and of high-grade disease at onset increase age, while risks of clinical diagnosis (i.e., diagnosis that would occur in the absence of PSA screening) and of metastasis (Figure B) increase with PSA levels.
The model allows two stages (local-regional or distant) and two grades (Gleason 2–7 or 8–10). Stage progression is permitted, but grade is fixed at the time of disease onset. To calibrate the model, we superimpose PSA screening2 and biopsy frequencies3 according to observed US screening patterns and obtain model-projected disease incidence. We use a simulated maximum likelihood algorithm based on a Poisson-type likelihood function to identify rates of onset, metastasis, and clinical diagnosis so that model-projected incidence matches incidence observed in the Surveillance, Epidemiology, and End Results (SEER) program by age, year, stage, and grade. The calibrated model closely replicates observed age-adjusted incidence rates by age, year, stage, and grade.4
The survival component of the model consists of disease-specific and other-cause survival. Disease-specific survival depends on age, stage (local-regional or distant), and grade (Gleason 2–7 or 8–10) at diagnosis. For local-regional cases, disease-specific survival also depends on primary treatment (radiation or surgery). In the absence of screening and treatment, disease-specific survival is based on data from cases diagnosed in SEER between 1983 and 1986, just prior to the PSA era. Treatment impacts survival after clinical diagnosis via a proportional hazards model with relative risks based on treatment trials5-7 and comparative effectiveness studies.8-10
Screening benefit arises from the detection of earlier stage tumors than might be identified without screening. For cases that are detected at an earlier stage by screening, their original survival from the time of clinical diagnosis is replaced by a survival time that corresponds to the earlier stage, leading to a reduction in the risk of prostate cancer death. We refer to this as a “stage-shift model” for the impact of screening on prostate cancer mortality.
Recently, the model was extended to accommodate dependence of the distribution of disease grade on age at onset. Additionally, the number of stages has been expanded to three (early localized, advanced localized, and distant) and the number of grades has been expanded to three (Gleason 2–6, 7, and 8–10). We also added several variations on the survival benefit model. In addition to a stage-shift model, we implemented a cure model, in which the benefit of screening is to cure a fraction of those cases who would have died of disease in the absence of screening, with the cure fraction constant, lead time-dependent, or PSA-dependent. We have also developed a flexible stage-shift model, in which the full improvement in survival that would be expected from a stage shift is downweighted using a parameter that is proportional to an individual’s lead time.11
References
- Thompson IM, Goodman PJ, Tangen CM, et al. The influence of finasteride on the development of prostate cancer. N. Engl. J. Med. 2003;349(3):215-224. [Abstract]
- Mariotto A, Etzioni R, Krapcho M, Feuer EJ. Reconstructing prostate-specific antigen (PSA) testing patterns among black and white men in the US from Medicare claims and the National Health Interview Survey. Cancer. 2007;109(9):1877-1886. [Abstract]
- Pinsky PF, Andriole GL, Kramer BS, Hayes RB, Prorok PC, Gohagan JK. Prostate biopsy following a positive screen in the Prostate, Lung, Colorectal and Ovarian cancer screening trial. J. Urol. 2005;173(3):746-750; discussion 750-751. [Abstract]
- Gulati R, Inoue L, Katcher J, Hazelton W, Etzioni R. Calibrating disease progression models using population data: a critical precursor to policy development in cancer control. Biostatistics. 2010;11(4):707-719. [Abstract]
- Bill-Axelson A, Holmberg L, Ruutu M, et al. Radical prostatectomy versus watchful waiting in early prostate cancer. N. Engl. J. Med. 2011;364(18):1708-1717. [Abstract]
- Wilt T, Brawer MK, Jones K, Barry M. Radical Prostatectomy versus Observation for Localized Prostate Cancer. N. Engl. J. Med. 2012;367(3):203-213. [Abstract]
- Bolla M, Gonzalez D, Warde P, et al. Improved survival in patients with locally advanced prostate cancer treated with radiotherapy and goserelin. N. Engl. J. Med. 1997;337(5):295-300. [Abstract]
- Hadley J, Yabroff KR, Barrett MJ, Penson DF, Saigal CS, Potosky AL. Comparative effectiveness of prostate cancer treatments: evaluating statistical adjustments for confounding in observational data. J. Natl. Cancer Inst. 2010;102(23):1780-1793. [Abstract]
- Cooperberg MR, Vickers AJ, Broering JM, Carroll PR. Comparative risk-adjusted mortality outcomes after primary surgery, radiotherapy, or androgen-deprivation therapy for localized prostate cancer. Cancer. 2010;116(22):5226-5234. [Abstract]
- Boorjian SA, Karnes RJ, Viterbo R, et al. Long-term survival after radical prostatectomy versus external-beam radiotherapy for patients with high-risk prostate cancer. Cancer. 2011;117(13):2883-2891. [Abstract]
- Gulati R, Tsodikov A, Etzioni R, et al. Expected population impacts of discontinued prostate-specific antigen screening. Cancer. 2014;120(22):3519-3526. [Abstract]