Decision analysis and Bayesian methods in clinical trials

Abstract Introduction: There is no standard of care for first-line therapy of low-grade FL. In US practice, the most common strategy is rituximab with combination chemotherapy. However, the optimal choice of regimen remains controversial; options include RCVP, RCHOP and R-Fludarabine-based chemotherapy (RFlu). Because data from randomized clinical trials are not available and unlikely to be generated in the future, we performed a decision analysis comparing RCVP, RCHOP, and RFlu as first-line therapy for FL. Methods: We constructed a Markov model of sequential first- and second-line therapy based on prescribing patterns in the US. The endpoint of the model was quality-adjusted time to tertiary referral for therapy such as RIT or autologous transplant (≥3rd line). A literature review was performed of the Medline database and international meeting abstracts. Clinical trials of both untreated and previously treated patients were systematically evaluated using explicit eligibility criteria. Data were extracted regarding response rates, treatment-related mortality, and progression-free survival (PFS). Weighted estimates were obtained using a fixed effects meta-analysis. The model also incorporated published data on heath state utilities, risk of anthracycline cardiotoxicity and fludarabine-related delayed cytopenias. Primary and sensitivity analyses were performed using TreeAge software. Results: The optimal treatment strategy consisted of RCHOP in first-line followed by RFlu in second-line (9.0 quality-adjusted life years; QALYs). Strategies containing RCVP in either first- or second-line were inferior (6.2–7.7 QALYs). The model was sensitive to first-line PFS of RCHOP and RFlu when these were varied over the range of estimates obtained from individual published trials. The model was robust in sensitivity analysis of most other parameters, including rate of delayed cytopenias after RFlu, anthracycline cardiotoxicity, and quality of life adjustments. Conclusions: Using decision analysis, the optimal first-line therapy for low-grade FL is RCHOP, followed by RFlu in second-line. This strategy maximizes quality-adjusted time to tertiary therapy. Use of RCVP does not improve overall quality-adjusted time relative to more intensive therapies.

Download Full-text

Bayesian methods for setting sample sizes and choosing allocation ratios in phase II clinical trials with time-to-event endpoints

Statistics in Medicine ◽

10.1002/sim.6426 ◽

2015 ◽

Vol 34 (11) ◽

pp. 1889-1903 ◽

Cited By ~ 4

Author(s):

Amy Cotterill ◽

John Whitehead

Keyword(s):

Clinical Trials ◽

Phase Ii ◽

Bayesian Methods ◽

Sample Sizes ◽

Time To Event ◽

Phase Ii Clinical Trials

Download Full-text

Bayesian survival analysis in clinical trials: What methods are used in practice?

Clinical Trials ◽

10.1177/1740774516673362 ◽

2016 ◽

Vol 14 (1) ◽

pp. 78-87 ◽

Cited By ~ 7

Author(s):

Caroline Brard ◽

Gwénaël Le Teuff ◽

Marie-Cécile Le Deley ◽

Lisa V Hampson

Keyword(s):

Clinical Trials ◽

Survival Analysis ◽

Bayesian Methods ◽

Regression Models ◽

Final Analysis ◽

Bayesian Regression ◽

Phase Iii ◽

Informative Priors ◽

Primary Analysis ◽

Survival Analyses

Background Bayesian statistics are an appealing alternative to the traditional frequentist approach to designing, analysing, and reporting of clinical trials, especially in rare diseases. Time-to-event endpoints are widely used in many medical fields. There are additional complexities to designing Bayesian survival trials which arise from the need to specify a model for the survival distribution. The objective of this article was to critically review the use and reporting of Bayesian methods in survival trials. Methods A systematic review of clinical trials using Bayesian survival analyses was performed through PubMed and Web of Science databases. This was complemented by a full text search of the online repositories of pre-selected journals. Cost-effectiveness, dose-finding studies, meta-analyses, and methodological papers using clinical trials were excluded. Results In total, 28 articles met the inclusion criteria, 25 were original reports of clinical trials and 3 were re-analyses of a clinical trial. Most trials were in oncology (n = 25), were randomised controlled (n = 21) phase III trials (n = 13), and half considered a rare disease (n = 13). Bayesian approaches were used for monitoring in 14 trials and for the final analysis only in 14 trials. In the latter case, Bayesian survival analyses were used for the primary analysis in four cases, for the secondary analysis in seven cases, and for the trial re-analysis in three cases. Overall, 12 articles reported fitting Bayesian regression models (semi-parametric, n = 3; parametric, n = 9). Prior distributions were often incompletely reported: 20 articles did not define the prior distribution used for the parameter of interest. Over half of the trials used only non-informative priors for monitoring and the final analysis (n = 12) when it was specified. Indeed, no articles fitting Bayesian regression models placed informative priors on the parameter of interest. The prior for the treatment effect was based on historical data in only four trials. Decision rules were pre-defined in eight cases when trials used Bayesian monitoring, and in only one case when trials adopted a Bayesian approach to the final analysis. Conclusion Few trials implemented a Bayesian survival analysis and few incorporated external data into priors. There is scope to improve the quality of reporting of Bayesian methods in survival trials. Extension of the Consolidated Standards of Reporting Trials statement for reporting Bayesian clinical trials is recommended.

Download Full-text