Screening Based on the Risk of Cancer Calculation From Bayesian Hierarchical Changepoint and Mixture Models of Longitudinal Markers

The standard approach to early detection of disease with a quantitative marker is to set a population-based fixed reference level for making further individual screening or referral decisions. For many types of disease, including prostate and ovarian cancer, additional information is contained in the subject-specific temporal behavior of the marker, which exhibits a characteristic alteration early in the course of the disease. In this article we derive a Bayesian approach to screening based on calculation of the posterior probability of disease given longitudinal marker levels. The method is motivated by a randomized ovarian cancer screening trial in the United Kingdom comprising 22,000 women screened over 4 years with an additional 5 years of follow-up on average. Levels of the antigen CA125 were recorded annually in the screened arm. CA125 profiles of cases and controls from the U.K. trial are modeled using hierarchical changepoint and mixture models, posterior distributions are calculated using Markov chain Monte Carlo methods, and the model is used to calculate the Bayesian posterior risk of having ovarian cancer given a new subject's single or multiple longitudinal CA125 levels. A screening strategy based on the risk calculation is then evaluated using data from an independent screening trial of 5,550 women performed in Sweden. A longitudinal CA125 screening strategy based on calculation of the risk of ovarian cancer is proposed. Simulations of a prospective trial using a strategy based on the risk calculated from longitudinal CA125 values indicate potentially large increases in sensitivity for a given specificity compared to the standard approach based on a fixed CA125 reference level for all subjects.

[1]  N. Metropolis,et al.  Equation of State Calculations by Fast Computing Machines , 1953, Resonance.

[2]  Marvin Zelen,et al.  On the theory of screening for chronic diseases , 1969 .

[3]  A. F. Smith.  A Bayesian approach to inference about a change-point in a sequence of random variables , 1975 .

[4]  J. Ware,et al.  On the use of repeated measurements in regression analysis with dichotomous responses. , 1979, Biometrics.

[5]  A. F. Smith,et al.  Straight Lines with a Change‐Point: A Bayesian Analysis of Some Renal Transplant Data , 1980 .

[6]  N. B. Booth,et al.  A Bayesian approach to retrospective identification of change-points , 1982 .

[7]  Philip C. Prorok.  IN THE DESIGN OF A REPETITIVE SCREENING PROGRAM , 1982 .

[8]  R. Bast,et al.  Immunopathologic characterization of a monoclonal antibody that recognizes common surface antigens of human ovarian tumors of serous, endometrioid, and clear cell types. , 1983, American journal of clinical pathology.

[9]  N. Day,et al.  Simplified models of screening for chronic disease: estimation procedures from mass screening programmes. , 1984, Biometrics.

[10]  K. F. Turkman,et al.  Optimal Screening Methods , 1989 .

[11]  G. Eklund,et al.  Prospective evaluation of serum CA 125 levels in a normal population, phase I: the specificities of single and serial determinations in testing for ovarian cancer. , 1990, Gynecologic oncology.

[12]  Adrian F. M. Smith,et al.  Sampling-Based Approaches to Calculating Marginal Densities , 1990 .

[13]  K. Gordon,et al.  Modeling and Monitoring Biomedical Time Series , 1990 .

[14]  O. Barndorff-Nielsen.  Modified signed log likelihood ratio , 1991 .

[15]  T S Lau,et al.  On dependent repeated screening tests. , 1991, Biometrics.

[16]  Adrian F. M. Smith,et al.  Hierarchical Bayesian Analysis of Changepoint Problems , 1992 .

[17]  John R. van Nagell,et al.  Ovarian cancer screening , 1993 .

[18]  Ultrasound screening for ovarian cancer. , 1993, Nursing standard (Royal College of Nursing (Great Britain) : 1987).

[19]  J. Grudzinskas,et al.  Prevalence screening for ovarian cancer in postmenopausal women by CA 125 measurement and ultrasonography. , 1993, BMJ.

[20]  S. Skates.  On secant approximations to cumulative distribution functions , 1993 .

[21]  D. Stephens.  Bayesian Retrospective Multiple‐Changepoint Identification , 1994 .

[22]  Simon G. Thompson,et al.  Optimal Sequential Screening Guidelines for Quantitative Risk Factors Measured with Error , 1995 .

[23]  P. Green.  Reversible jump Markov chain Monte Carlo computation and Bayesian model determination , 1995 .

[24]  Clinical interest of estimation of the growth kinetics of ovarian primary recurrences by CA125 doubling time , 1997 .

[25]  H. D. de Bruijn,et al.  Effect of low-dose oral etoposide on serum CA-125 in patients with advanced epithelial ovarian cancer. , 1997, Gynecologic oncology.

[26]  Kathleen A. Cronin,et al.  Changepoint Modeling of Longitudinal PSA as a Biomarker for Prostate Cancer , 1997 .

[27]  I. Barillot,et al.  [Clinical value of the estimation of growth kinetics of primary ovarian cancer recurrences by CA125 doubling time]. , 1997, Bulletin du cancer.

[28]  Larry C. Clark,et al.  Using PSA to detect prostate cancer onset: An application of Bayesian retrospective and prospective changepoint identification , 1999 .