Risk Prediction for Epithelial Ovarian Cancer in 11 United States-Based Case-Control Studies: Incorporation of Epidemiologic Risk Factors and 17 Confirmed Genetic Loci

, on behalf of the Ovarian Cancer Association Consortium

Research output: Contribution to journalArticle

Abstract

Previously developed models for predicting absolute risk of invasive epithelial ovarian cancer have included a limited number of risk factors and have had low discriminatory power (area under the receiver operating characteristic curve (AUC) < 0.60). Because of this, we developed and internally validated a relative risk prediction model that incorporates 17 established epidemiologic risk factors and 17 genome-wide significant single nucleotide polymorphisms (SNPs) using data from 11 case-control studies in the United States (5,793 cases; 9,512 controls) from the Ovarian Cancer Association Consortium (data accrued from 1992 to 2010). We developed a hierarchical logistic regression model for predicting case-control status that included imputation of missing data. We randomly divided the data into an 80% training sample and used the remaining 20% for model evaluation. The AUC for the full model was 0.664. A reduced model without SNPs performed similarly (AUC = 0.649). Both models performed better than a baseline model that included age and study site only (AUC = 0.563). The best predictive power was obtained in the full model among women younger than 50 years of age (AUC = 0.714); however, the addition of SNPs increased the AUC the most for women older than 50 years of age (AUC = 0.638 vs. 0.616). Adapting this improved model to estimate absolute risk and evaluating it in prospective data sets is warranted.

Original languageEnglish (US)
Pages (from-to)579-589
Number of pages11
JournalAmerican Journal of Epidemiology
Volume184
Issue number8
StatePublished - Oct 15 2016

Fingerprint

Epidemiologic Factors
Genetic Loci
Area Under Curve
Case-Control Studies
Single Nucleotide Polymorphism
Logistic Models
Ovarian epithelial cancer
ROC Curve
Ovarian Neoplasms
Genome

Keywords

  • genetic risk polymorphisms
  • model evaluation
  • ovarian cancer
  • risk model

ASJC Scopus subject areas

  • Epidemiology

Cite this

Risk Prediction for Epithelial Ovarian Cancer in 11 United States-Based Case-Control Studies : Incorporation of Epidemiologic Risk Factors and 17 Confirmed Genetic Loci. / , on behalf of the Ovarian Cancer Association Consortium.

In: American Journal of Epidemiology, Vol. 184, No. 8, 15.10.2016, p. 579-589.

Research output: Contribution to journalArticle

@article{2782ae097d324ed0a239c661beada346,
title = "Risk Prediction for Epithelial Ovarian Cancer in 11 United States-Based Case-Control Studies: Incorporation of Epidemiologic Risk Factors and 17 Confirmed Genetic Loci",
abstract = "Previously developed models for predicting absolute risk of invasive epithelial ovarian cancer have included a limited number of risk factors and have had low discriminatory power (area under the receiver operating characteristic curve (AUC) < 0.60). Because of this, we developed and internally validated a relative risk prediction model that incorporates 17 established epidemiologic risk factors and 17 genome-wide significant single nucleotide polymorphisms (SNPs) using data from 11 case-control studies in the United States (5,793 cases; 9,512 controls) from the Ovarian Cancer Association Consortium (data accrued from 1992 to 2010). We developed a hierarchical logistic regression model for predicting case-control status that included imputation of missing data. We randomly divided the data into an 80{\%} training sample and used the remaining 20{\%} for model evaluation. The AUC for the full model was 0.664. A reduced model without SNPs performed similarly (AUC = 0.649). Both models performed better than a baseline model that included age and study site only (AUC = 0.563). The best predictive power was obtained in the full model among women younger than 50 years of age (AUC = 0.714); however, the addition of SNPs increased the AUC the most for women older than 50 years of age (AUC = 0.638 vs. 0.616). Adapting this improved model to estimate absolute risk and evaluating it in prospective data sets is warranted.",
keywords = "genetic risk polymorphisms, model evaluation, ovarian cancer, risk model",
author = "{, on behalf of the Ovarian Cancer Association Consortium} and Clyde, {Merlise A.} and {Palmieri Weber}, Rachel and Iversen, {Edwin S.} and Poole, {Elizabeth M.} and Doherty, {Jennifer A.} and Goodman, {Marc T.} and Ness, {Roberta B.} and Risch, {Harvey A.} and Rossing, {Mary Anne} and Terry, {Kathryn L.} and Nicolas Wentzensen and Whittemore, {Alice S.} and Hoda Anton-Culver and Bandera, {Elisa V.} and Andrew Berchuck and Carney, {Michael E.} and Cramer, {Daniel W.} and Cunningham, {Julie M.} and Cushing-Haugen, {Kara L.} and Edwards, {Robert P.} and Fridley, {Brooke L.} and Goode, {Ellen L.} and Galina Lurie and Valerie McGuire and Francesmary Modugno and Moysich, {Kirsten B.} and Olson, {Sara H.} and Pearce, {Celeste Leigh} and Pike, {Malcolm C.} and Rothstein, {Joseph H.} and Sellers, {Thomas A.} and Weiva Sieh and Daniel Stram and Thompson, {Pamela J.} and Vierkant, {Robert A.} and Wicklund, {Kristine G.} and Wu, {Anna H.} and Argyrios Ziogas and Tworoger, {Shelley S.} and Schildkraut, {Joellen M.}",
year = "2016",
month = "10",
day = "15",
language = "English (US)",
volume = "184",
pages = "579--589",
journal = "American Journal of Epidemiology",
issn = "0002-9262",
publisher = "Oxford University Press",
number = "8",

}

TY - JOUR

T1 - Risk Prediction for Epithelial Ovarian Cancer in 11 United States-Based Case-Control Studies

T2 - Incorporation of Epidemiologic Risk Factors and 17 Confirmed Genetic Loci

AU - , on behalf of the Ovarian Cancer Association Consortium

AU - Clyde, Merlise A.

AU - Palmieri Weber, Rachel

AU - Iversen, Edwin S.

AU - Poole, Elizabeth M.

AU - Doherty, Jennifer A.

AU - Goodman, Marc T.

AU - Ness, Roberta B.

AU - Risch, Harvey A.

AU - Rossing, Mary Anne

AU - Terry, Kathryn L.

AU - Wentzensen, Nicolas

AU - Whittemore, Alice S.

AU - Anton-Culver, Hoda

AU - Bandera, Elisa V.

AU - Berchuck, Andrew

AU - Carney, Michael E.

AU - Cramer, Daniel W.

AU - Cunningham, Julie M.

AU - Cushing-Haugen, Kara L.

AU - Edwards, Robert P.

AU - Fridley, Brooke L.

AU - Goode, Ellen L.

AU - Lurie, Galina

AU - McGuire, Valerie

AU - Modugno, Francesmary

AU - Moysich, Kirsten B.

AU - Olson, Sara H.

AU - Pearce, Celeste Leigh

AU - Pike, Malcolm C.

AU - Rothstein, Joseph H.

AU - Sellers, Thomas A.

AU - Sieh, Weiva

AU - Stram, Daniel

AU - Thompson, Pamela J.

AU - Vierkant, Robert A.

AU - Wicklund, Kristine G.

AU - Wu, Anna H.

AU - Ziogas, Argyrios

AU - Tworoger, Shelley S.

AU - Schildkraut, Joellen M.

PY - 2016/10/15

Y1 - 2016/10/15

N2 - Previously developed models for predicting absolute risk of invasive epithelial ovarian cancer have included a limited number of risk factors and have had low discriminatory power (area under the receiver operating characteristic curve (AUC) < 0.60). Because of this, we developed and internally validated a relative risk prediction model that incorporates 17 established epidemiologic risk factors and 17 genome-wide significant single nucleotide polymorphisms (SNPs) using data from 11 case-control studies in the United States (5,793 cases; 9,512 controls) from the Ovarian Cancer Association Consortium (data accrued from 1992 to 2010). We developed a hierarchical logistic regression model for predicting case-control status that included imputation of missing data. We randomly divided the data into an 80% training sample and used the remaining 20% for model evaluation. The AUC for the full model was 0.664. A reduced model without SNPs performed similarly (AUC = 0.649). Both models performed better than a baseline model that included age and study site only (AUC = 0.563). The best predictive power was obtained in the full model among women younger than 50 years of age (AUC = 0.714); however, the addition of SNPs increased the AUC the most for women older than 50 years of age (AUC = 0.638 vs. 0.616). Adapting this improved model to estimate absolute risk and evaluating it in prospective data sets is warranted.

AB - Previously developed models for predicting absolute risk of invasive epithelial ovarian cancer have included a limited number of risk factors and have had low discriminatory power (area under the receiver operating characteristic curve (AUC) < 0.60). Because of this, we developed and internally validated a relative risk prediction model that incorporates 17 established epidemiologic risk factors and 17 genome-wide significant single nucleotide polymorphisms (SNPs) using data from 11 case-control studies in the United States (5,793 cases; 9,512 controls) from the Ovarian Cancer Association Consortium (data accrued from 1992 to 2010). We developed a hierarchical logistic regression model for predicting case-control status that included imputation of missing data. We randomly divided the data into an 80% training sample and used the remaining 20% for model evaluation. The AUC for the full model was 0.664. A reduced model without SNPs performed similarly (AUC = 0.649). Both models performed better than a baseline model that included age and study site only (AUC = 0.563). The best predictive power was obtained in the full model among women younger than 50 years of age (AUC = 0.714); however, the addition of SNPs increased the AUC the most for women older than 50 years of age (AUC = 0.638 vs. 0.616). Adapting this improved model to estimate absolute risk and evaluating it in prospective data sets is warranted.

KW - genetic risk polymorphisms

KW - model evaluation

KW - ovarian cancer

KW - risk model

UR - http://www.scopus.com/inward/record.url?scp=85021857336&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85021857336&partnerID=8YFLogxK

M3 - Article

C2 - 27698005

VL - 184

SP - 579

EP - 589

JO - American Journal of Epidemiology

JF - American Journal of Epidemiology

SN - 0002-9262

IS - 8

ER -