A new method for synthesizing radiation dose-response data from multiple trials applied to prostate cancer

Area of Science:

• Oncology
• Biostatistics
• Clinical Trial Design

Background:

• Accurate dose-response data is crucial for optimizing prostate cancer treatment.
• Synthesizing data across studies requires robust analytical methods to account for design differences.

Purpose of the Study:

• To develop a novel method for synthesizing dose-response data for prostate cancer biochemical control.
• To compare dose-response gradients between randomized and nonrandomized studies and across risk groups.

Main Methods:

• Reviewed nine prostate cancer dose escalation studies (6,539 patients) from MEDLINE and CINAHL.
• Estimated the normalized dose-response gradient, gamma(50), for each study.
• Synthesized gamma(50) across studies using a novel analytical approach.

Main Results:

• Nonrandomized studies showed a significantly higher gamma(50) than randomized studies for intermediate-high risk patients (1.63 vs. 0.93, p=0.03).
• A borderline significant difference was observed for low-risk patients (1.78 vs. 0.56, p=0.08).
• Pooled data from randomized trials yielded an overall gamma(50) of 0.84 (95% CI: 0.54-1.15).

Conclusions:

• Nonrandomized studies tend to overestimate the steepness of the dose-response curve compared to randomized trials.
• Potential reasons for overestimation include stage migration and differences in follow-up.
• This overestimation can lead to inflated expectations for dose-escalation and potentially underpowered clinical trials.