Aline Talhouk

PHD

Department of Obstetrics and Gynecology, University of British Columbia, Vancouver, Canada

Assistant Professor

PLCO (Learn more about this study)

PLCO-878

Dec 13, 2021

Reproducibility and external validation of epidemiological risk models for endometrial cancer

Endometrial cancer (EC), or cancer of the uterus, is the most common gynecological malignancy in the developed world, and the second most common cancer overall in women with an intact uterus (1). Unlike most other cancers, annual incidence of EC and mortality from this cancer has been increasing worldwide, particularly in developed countries (2–4). The incidence rate of EC is projected to continue to rise over the next decade and mortality rates are expected to double (5–7). This rapid increase is believed to be driven by the increasing prevalence of obesity and an aging population, both known as risk factors for EC. A recent priority setting exercise with EC patients identified personalized risk prediction to be at the forefront of both patients and clinicians' priorities (8). Identifying asymptomatic women at high risk of EC, in the population, would allow targeted prevention strategies and directed screening (9) to enable early detection.

Despite EC being the most common gynecological malignancy, only a few population-based epidemiological models have been proposed to predict a woman’s absolute risk of an EC diagnosis using risk factors (10). The first model was developed by Pfeiffer et al (11) in 2013 from two large prospective population-based United States cohorts (Prostate, Lung, Colorectal, and Ovarian Cancer (PLCO) and NIH-AARP cohorts) of non-hispanic, white women aged 50+. The second model was developed by Hüsing et al (12) in 2015 using a large dataset of women without hysterectomy or prevalent cancer from a European Prospective Investigation into Cancer and Nutrition cohort (EPIC). Both models resulted in a prediction of the individual’s absolute risk of developing invasive EC. Recently, a machine learning model was proposed using the PLCO cancer screening trial data (Hart et al (13)). Both models by Pfeiffer and Hüsing et al, that used traditional epidemiological approaches scored in internal validation an AUC of 0.68 and 0.77 respectively, while the Hart model achieved a much higher AUC of 0.96. These models if externally validated to accurately predict an individual’s absolute risk could potentially help identify populations that benefit from prophylactic interventions or who can be directed to screening for early detection of cancer.

. Study purpose:
Our team has recently obtained funding to validate these three models in a Canadian population (CanPath). Furthermore, we want to compare the performance of these models in datasets that were not used to derive them. The Pfeiffer and Husing models are reproducible from the publications, however, the Hart model isn’t. In this study, we would like to request the PLCO data to reproduce the Hart model (13) and to validate the Hüsing model (12).
. Specific aims:
Using the PLCO dataset, we plan to undertake a study to achieve the following two aims:
Aim 1: Reproduce Hart model parameters so that we can validate it in an independent cohort.
Aim 2: Evaluate the Husing model on the PLCO data.

References
1. Mullins MA, Cote ML. Beyond Obesity: The Rising Incidence and Mortality Rates of Uterine Corpus Cancer. J Clin Oncol. 2019;37(22):1851-1853. doi:10.1200/JCO.19.01240
2. Canadian Cancer Statistics publication -Canadian Cancer Society. cancer.ca/Canadian-Cancer-Statistics-2017-EN.pdf. Published 2017. Accessed February 22, 2018.
3. Lortet-Tieulent J, Ferlay J, Bray F, Jemal A. International Patterns and Trends in Endometrial Cancer Incidence, 1978–2013. JNCI J Natl Cancer Inst. 2018;110(4):354-361. doi:10.1093/jnci/djx214
4. Evans T, Sany O, Pearmain P, Ganesan R, Blann A, Sundar S. Differential trends in the rising incidence of endometrial cancer by type: data from a UK population-based registry from 1994 to 2006. Br J Cancer. 2011;104(9):1505-1510. doi:10.1038/bjc.2011.68
5. Sheikh MA, Althouse AD, Freese KE, et al. USA Endometrial Cancer Projections to 2030: should we be concerned? Futur Oncol. 2014;10(16):2561-2568. doi:10.2217/fon.14.192
6. Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM, Matrisian LM. Projecting Cancer Incidence and Deaths to 2030: The Unexpected Burden of Thyroid, Liver, and Pancreas Cancers in the United States. Cancer Res. 2014;74(11):2913-2921. doi:10.1158/0008-5472.CAN-14-0155
7. Zhuo Z, Wang A, Yu H. Metformin targeting autophagy overcomes progesterone resistance in endometrial carcinoma. Arch Gynecol Obstet. 2016;294(5):1055-1061. doi:10.1007/s00404-016-4148-0
8. Wan YL, Beverley-Stevenson R, Carlisle D, et al. Working together to shape the endometrial cancer research agenda:The top ten unanswered research questions. Gynecol Oncol. 2016;143(2):287-293. doi:10.1016/j.ygyno.2016.08.333
9. Costas L, Frias-Gomez J, Guardiola M, et al. New perspectives on screening and early detection of endometrial cancer. Int J Cancer. June 2019:ijc.32514. doi:10.1002/ijc.32514
10. Alblas M, Velt KB, Pashayan N, Widschwendter M, Steyerberg EW, Vergouwe Y. Prediction models for endometrial cancer for the general population or symptomatic women: A systematic review. Crit Rev Oncol Hematol. 2018;126:92-99. doi:10.1016/J.CRITREVONC.2018.03.023
11. Pfeiffer RM, Park Y, Kreimer AR, et al. Risk Prediction for Breast, Endometrial, and Ovarian Cancer in White Women Aged 50 y or Older: Derivation and Validation from Population-Based Cohort Studies. Franco EL, ed. PLoS Med. 2013;10(7):e1001492. doi:10.1371/journal.pmed.1001492
12. Hüsing A, Dossus L, Ferrari P, et al. An epidemiological model for prediction of endometrial cancer risk in Europe. Eur J Epidemiol. 2016;31(1):51-60. doi:10.1007/s10654-015-0030-9
13. Hart GR, Yan V, Huang GS, et al. Population-Based Screening for Endometrial Cancer: Human vs. Machine Intelligence. Front Artif Intell. 2020; 3:539879. doi: 10.3389/frai.2020.539879. eCollection 2020.

1. Gillian Hanley
Email: gillian.hanley@vch.ca
Role: Epidemiologist

2. Lecuyer Mathias
Email: mathias.lecuyer@ubc.ca
Role: Computer Scientist

3. Chiu Derek
Email: dchiu@bccrc.ca
Role: Statistical Analyst