SCOPE (Michigan) University Of Michigan

Overview

The SCOPE model preserves key components of the University of Michigan Self-Consistency Analysis of Surveillance (SCANS) analytic model while extending the representation of preclinical disease through longitudinal PSA dynamics and an underlying latent tumor growth process.

The SCANS model is an analytic model of prostate cancer incidence, diagnosis, treatment, and survival comprised of three components.¹

(1) A model for age at disease onset and age at clinical diagnosis, where the interval between these events defines the sojourn time. Onset age for each person is defined as the youngest age at which prostate cancer could potentially be detected by a biopsy. Screening based on PSA tests can lead to earlier, screen-detected diagnosis, with advancement in time relative to clinical diagnosis defined as the lead time.
(2) A model for disease characteristics at diagnosis (stage and grade), specified using a multinomial framework that depends on mode of detection and lead time.^2,3
(3) A model for post-diagnosis treatment and survival, conditional on stage, grade, and lead time.

The SCANS model was estimated based on maximum likelihood and algorithms specifically developed to make the comprehensive joint model tractable using population-based incidence and survival data from SEER.^4,5 It has been used to study prostate cancer trends and screening outcomes.^6-9

SCOPE preserves the SCANS formulations for age at onset and post-diagnosis survival and is calibrated to reproduce key SCANS targets, including the distribution of sojourn times and the joint distribution of stage and grade at diagnosis. However, rather than representing preclinical disease implicitly, SCOPE introduces an explicit latent tumor growth process that generates longitudinal PSA trajectories, disease progression, and clinical detection at the individual level. This framework allows representation of longitudinal biomarkers, dynamic screening decisions, and individual disease trajectories while maintaining consistency with the established SCANS framework.

The model can be used to evaluate comparative effectiveness and cost-effectiveness of prostate cancer screening and treatment strategies, including risk-based approaches that tailor screening intervals, biopsy decisions, and stopping ages based on individual risk profiles. By incorporating an underlying latent tumor growth process, the model provides a unified framework in which longitudinal PSA trajectories, stage progression, grade progression, and clinical detection arise from a common underlying disease process. The current implementation allows diagnostic test performance to depend on disease characteristics such as grade and provides a flexible platform for evaluating dynamic, personalized screening and diagnostic pathways.

References

1. Tsodikov A, Szabo A, Wegelin J. A population model of prostate cancer incidence. Statistics in Medicine. Wiley Online Library; 2006;25(16):2846–2866. [Abstract]
2. Tsodikov A, Chefo S. Generalized self-consistency: Multinomial logit model and Poisson likelihood. Journal of statistical planning and inference. Elsevier; 2008;138(8):2380–2397. [Abstract]
3. Chefo S, Tsodikov A. Stage-specific cancer incidence: An artificially mixed multinomial logit model. Statistics in medicine. Wiley Online Library; 2009;28(15):2054–2076. [Abstract]
4. Wang S, Tsodikov A. A self-consistency approach to multinomial logit model with random effects. Journal of statistical planning and inference. Elsevier; 2010;140(7):1939–1947.
5. Tsodikov A, Liu LX, Tseng C. Likelihood transformations and artificial mixtures. Statistical Modeling for Biological Systems: In Memory of Andrei Yakovlev. Springer; 2020. p. 191–209.
6. Tsodikov A, Gulati R, Heijnsdijk EA, Pinsky PF, Moss SM, Qiu S, De Carvalho TM, Hugosson J, Berg CD, Auvinen A, others. Reconciling the effects of screening on prostate cancer mortality in the ERSPC and PLCO trials. Annals of internal medicine. American College of Physicians; 2017;167(7):449–455. [Abstract]
7. Tsodikov A, Gulati R, de Carvalho TM, Heijnsdijk EA, Hunter-Merrill RA, Mariotto AB, de Koning HJ, Etzioni R. Is prostate cancer different in black men? Answers from 3 natural history models. Cancer. Wiley Online Library; 2017;123(12):2312–2319. [Abstract]
8. Gulati R, Tsodikov A, Etzioni R, Hunter-Merrill RA, Gore JL, Mariotto AB, Cooperberg MR. Expected population impacts of discontinued prostate-specific antigen screening. Cancer. Wiley Online Library; 2014;120(22):3519–3526. [Abstract]
9. Etzioni R, Tsodikov A, Mariotto A, Szabo A, Falcon S, Wegelin J, Ditommaso D, Karnofski K, Gulati R, Penson DF, others. Quantifying the role of PSA screening in the US prostate cancer mortality decline. Cancer Causes & Control. Springer; 2008;19(2):175–181. [Abstract]