Risk bimarkers have become increasingly important in clinical decision making, guiding patients and their clinicians in choosing the most appropriate course of therapy or surveillance after treatment. Constructing accurate and individualized prediction rules and conducting rigorous validation are critical to the cancer biomarker field. Prospective cohort studies are crucial for such evaluation as time to event carries more information about a marker's value on early detection and prognosis than a simple measure of disease status. But prospective biomarker evaluation is challenging. Until now there has been little guidance pro-vided for statistical design and analysis of these studies. We propose to extend our previously funded effort to address several new challenges in prospective marker evaluation. The proposal will emphasize three unique aspects in prospective biomarker evaluation. First, for many cancers disease outcome may be heterogeneous due to the biological nature of the disease or selection of treatments. Constructing and validating prognostic and treatment selection rules based on more specific prediction of the risk of develop-ing aggressive cancer as opposed to indolent cancer is of great clinical interest yet analytically challenging.
In Aim 1 we will provide statistical tools for developing and validating risk markers in a population with an unknown mixture of indolent and aggressive cancers. We propose statistical methods that facilitate the development and evaluation of prognostic markers for risk stratification. Methods for deriving and evalu-ating individualized treatment rules in the presence of a mixture of indolent and aggressive cancers will be considered. Second, among patients diagnosed with cancer who chose to be on active surveillance, developing monitoring tools to make adaptive monitoring or intervention recommendations with longitudinal biomarkers may alleviate overtreatment without missing signs of progression.
In Aim 2 we will consider flexible procedures to quantify the updated predictive accuracy of longitudinal markers. In addition, we will develop and evaluate decision rules on the basis of risk, incorporating both cross-sectional and longi-tudinal marker information. The ascertainment of marker information in a large cohort requires enormous resources. Cost-effective cohort sampling is therefore highly desirable.
In Aim 3 we will develop procedures to improve the efficiency of estimating risk and accuracy parameters and rigorously evaluate and compare different choices of matching/stratification rules and identify optimal pairs of analyses and sampling strate-gies. We will also develop estimation procedures for evaluating longitudinal markers in two-phase studies. Applications in cancer biomarker development provide a context for our research. Data from the Early De-tection Research Network and from several large cohort studies will be analyzed. Programs and algorithms developed in this proposal will be made available to public.

Public Health Relevance

The research proposal addresses the pressing need for strong statistical input in the cancer biomarker field, providing comprehensive statistical tools that will enable investigators to conduct valid and more powerful biomarker validation studies and to evaluate the prognostic and treatment-selection potential of novel biomarkers. Integrating our research into clinical settings will help improve survival outcomes and reduce the burden of cancer treatment.

Agency
National Institute of Health (NIH)
Institute
National Institute of General Medical Sciences (NIGMS)
Type
Research Project (R01)
Project #
5R01GM085047-07
Application #
9058556
Study Section
Cancer Biomarkers Study Section (CBSS)
Program Officer
Marcus, Stephen
Project Start
2009-09-04
Project End
2018-04-30
Budget Start
2016-05-01
Budget End
2017-04-30
Support Year
7
Fiscal Year
2016
Total Cost
Indirect Cost
Name
Fred Hutchinson Cancer Research Center
Department
Type
DUNS #
078200995
City
Seattle
State
WA
Country
United States
Zip Code
98109
Liu, Dandan; Cai, Tianxi; Lok, Anna et al. (2018) Nonparametric Maximum Likelihood Estimators of Time-Dependent Accuracy Measures for Survival Outcome Under Two-Stage Sampling Designs. J Am Stat Assoc 113:882-892
Kearns, James T; Faino, Anna V; Newcomb, Lisa F et al. (2018) Role of Surveillance Biopsy with No Cancer as a Prognostic Marker for Reclassification: Results from the Canary Prostate Active Surveillance Study. Eur Urol 73:706-712
Cooperberg, Matthew R; Brooks, James D; Faino, Anna V et al. (2018) Refined Analysis of Prostate-specific Antigen Kinetics to Predict Prostate Cancer Active Surveillance Outcomes. Eur Urol 74:211-217
Lin, Daniel W; Newcomb, Lisa F; Brown, Marshall D et al. (2017) Evaluating the Four Kallikrein Panel of the 4Kscore for Prediction of High-grade Prostate Cancer in Men in the Canary Prostate Active Surveillance Study. Eur Urol 72:448-454
Diamandis, Eleftherios P; Eklund, Emma; Muytjens, Carla et al. (2017) Effect of age on serum prostate-specific antigen in women. Clin Chem Lab Med 55:e271-e272
Diamandis, Eleftherios P; Stanczyk, Frank Z; Wheeler, Sarah et al. (2017) Serum complexed and free prostate-specific antigen (PSA) for the diagnosis of the polycystic ovarian syndrome (PCOS). Clin Chem Lab Med 55:1789-1797
Zheng, Yingye; Brown, Marshall; Lok, Anna et al. (2017) IMPROVING EFFICIENCY IN BIOMARKER INCREMENTAL VALUE EVALUATION UNDER TWO-PHASE DESIGNS. Ann Appl Stat 11:638-654
Maziarz, Marlena; Heagerty, Patrick; Cai, Tianxi et al. (2017) On longitudinal prediction with time-to-event outcome: Comparison of modeling options. Biometrics 73:83-93
Zhou, Qian M; Dai, Wei; Zheng, Yingye et al. (2017) Robust Dynamic Risk Prediction with Longitudinal Studies. Stat Theory Relat Fields 1:159-170
Macleod, Liam C; Ellis, William J; Newcomb, Lisa F et al. (2017) Timing of Adverse Prostate Cancer Reclassification on First Surveillance Biopsy: Results from the Canary Prostate Cancer Active Surveillance Study. J Urol 197:1026-1033

Showing the most recent 10 out of 29 publications