A two-stage model in a Bayesian framework to estimate a survival endpoint in the presence of confounding by indication

Abstract

Background

Biomarker series can indicate disease progression and predict clinical endpoints. When a treatment is prescribed depending on the biomarker, confounding by indication might be introduced if the treatment modifies the marker profile and risk of failure.

Objective

Our aim was to highlight the flexibility of a two-stage model fitted within a Bayesian Markov Chain Monte Carlo framework. For this purpose, we monitored the prostate-specific antigens in prostate cancer patients treated with external beam radiation therapy. In the presence of rising prostate-specific antigens after external beam radiation therapy, salvage hormone therapy can be prescribed to reduce both the prostate-specific antigens concentration and the risk of clinical failure, an illustration of confounding by indication. We focused on the assessment of the prognostic value of hormone therapy and prostate-specific antigens trajectory on the risk of failure.

Methods

We used a two-stage model within a Bayesian framework to assess the role of the prostate-specific antigens profile on clinical failure while accounting for a secondary treatment prescribed by indication. We modeled prostate-specific antigens using a hierarchical piecewise linear trajectory with a random changepoint. Residual prostate-specific antigens variability was expressed as a function of prostate-specific antigens concentration. Covariates in the survival model included hormone therapy, baseline characteristics, and individual predictions of the prostate-specific antigens nadir and timing and prostate-specific antigens slopes before and after the nadir as provided by the longitudinal process.

Results

We showed positive associations between an increased prostate-specific antigens nadir, an earlier changepoint and a steeper post-nadir slope with an increased risk of failure. Importantly, we highlighted a significant benefit of hormone therapy, an effect that was not observed when the prostate-specific antigens trajectory was not accounted for in the survival model.

Conclusion

Our modeling strategy was particularly flexible and accounted for multiple complex features of longitudinal and survival data, including the presence of a random changepoint and a time-dependent covariate.

Keywords

Two-stage model Bayesian analysis hierarchical model changepoint simulations MCMC prostate cancer PSA

Get full access to this article

View all access options for this article.

References

Robins

. The control of confounding by intermediate variables. Stat Med 1989; 8: 679–701.

Kennedy

Taylor

JMG

Schaubel

et al.

The effect of salvage therapy on survival in a longitudinal study with confounding by indication. Stat Med 2010; 29: 2569–2580.

Eastham

Riedel

Scardino

et al.

Variation of serum prostate-specific antigen levels. J Am Med Assoc 2003; 289: 2695–2700.

Ballentine Carter

Pearson

Metter

et al.

Longitudinal evaluation of prostate-specific antigen levels in men with and without prostate disease. J Am Med Assoc 1992; 267: 2215–2220.

Crook

Bahadur

Robertson

et al.

Evaluation of radiation effect, tumor differentiation, and prostate specific antigen staining in sequential prostate biopsies after external beam radiotherapy for patients with prostate carcinoma. Cancer 1997; 79: 81–89.

Crook

Choan

Perry

et al.

Serum prostate-specific antigen profile following radiotherapy for prostate cancer: implications for patterns of failure and definition of cure. Urology 1998; 51: 566–572.

Albertsen

Hanley

Penson

et al.

Validation of increasing prostate specific antigen as a predictor of prostate cancer death after treatment of localized prostate cancer with surgery or radiation. J Urol 2004; 171: 2221–2225.

Valicenti

DeSilvio

Hanks

et al.

Posttreatment prostatic-specific antigen doubling time as a surrogate endpoint for prostate cancer-specific survival: an analysis of Radiation Therapy Oncology Group Protocol 92-02. Int J Radiat Oncol Biol Phys 2006; 66: 1064–1071.

Ray

Thames

Levy

et al.

PSA nadir predicts biochemical and distant failures after external beam radiotherapy for prostate cancer: a multi-institutional analysis. Int J Radiat Oncol Biol Phys 2006; 64: 1140–1150.

10.

Cheung

Tucker

Kuban

. First-year PSA kinetics and minima after prostate cancer radiotherapy are predictive of overall survival. Int J Radiat Oncol Biol Phys 2006; 66: 20–24.

11.

Hanlon

Diratzouian

Hanks

. Posttreatment prostate-specific antigen nadir highly predictive of distant failure and death from prostate cancer. Int J Radiat Oncol Biol Phys 2002; 53: 297–303.

12.

Tsiatis

DeGruttola

Wulfsohn

. Modeling the relationship of survival to longitudinal data measured with error: applications to survival and CD4 counts in patients with AIDS. J Am Stat Assoc 1995; 90: 27–37.

13.

Bellera

Hanley

Joseph

et al.

Detecting trends in noisy data series: application to biomarker series. Am J Epidemiol 2008; 167: 1130–1139.

14.

Taylor

JMG

Sandler

. Individualized predictions of disease progression following radiotherapy for prostate cancer. J Clin Oncol 2005; 23: 816–825.

15.

Michalski

Winter

Purdy

et al.

Toxicity after three-dimensional radiotherapy for prostate cancer on RTOG 9406 dose Level V. Int J Radiat Oncol Biol Phys 2005; 62: 706–713.

16.

Roach

Winter

Michalski

et al.

Penile bulb dose and impotence after three-dimensional conformal radiotherapy for prostate cancer on RTOG 9406: findings from a prospective, multi-institutional, Phase I/II dose-escalation study. Int J Radiat Oncol Biol Phys 2004; 60: 1351–1356.

17.

Kestin

Vicini

Ziaja

et al.

Defining biochemical cure for prostate carcinoma patients treated with external beam radiation therapy. Cancer 1999; 86: 1557–1566.

18.

Proust-Lima

Taylor

JMG

Williams

et al.

Determinants of change in prostate-specific antigen over time and its association with recurrence after external beam radiation therapy for prostate cancer in five large cohorts. Int J Radiat Oncol Biol Phys 2008; 72: 782–791.

19.

Schmid

McNeal

Stamey

. Observations on the doubling time of prostate cancer. The use of serial prostate-specific antigen in patients with untreated disease as a measure of increasing volume. Cancer 1993; 71: 2031–2040.

20.

Meek

Park

Oberman

et al.

A prospective study of prostate specific antigen levels in patients receiving radiotherapy for localized carcinoma of the prostate. Int J Radiat Oncol Biol Phys 1990; 19: 733–741.

21.

Slate

Cronin

Changepoint modeling of longitudinal PSA as biomarker for prostate cancer. In: Gatsonis

Hodges

Kass

et al. (eds). Case studies in Bayesian statistics 1999; Vol 3, New York: Springer-Verlag New York, Inc, pp. 435–456. .

22.

Gelman

. Prior distributions for variance parameters in hierarchical models. Bayesian Anal 2006; 1: 515–533.

23.

Spiegelhalter

Thomas

Best

et al.

WinBUGS: Bayesian inference using Gibbs sampling, version 1.3, Cambridge, UK: MRC Biostatistics Unit, 2000.

24.

Gelman

Carlin

Stern

et al.

Bayesian data analysis, 2nd ed. Boca Raton, FL: CRC Press, 2004. .

25.

Raftery

Lewis

How many iterations in the Gibbs sampler? In: Bernardo

Dawid

(eds). Bayesian statistics 1992; Vol 4, Oxford: Oxford University Press, pp. 763–773. .

26.

Taylor

JMG

Shen

Kennedy

et al.

Comparison of methods for estimating the effect of slavage therapy in prostate cancer when treatment is given by indication. Stat Med 2014; 33: 257–274.

27.

Wulfsohn

Tsiatis

. A joint model for survival and longitudinal data measured with error. Biometrics 1997; 53: 330–339.

28.

Sweeting

Thompson

. Joint modelling of longitudinal and time-to-event data with application to predicting abnominal aortic aneurysm growth and rupture. Biometric J 2011; 53: 750–763.