New Paper Alert! Addressing Postprogression Therapy Confounding in Phase III Oncology Trials: A Crucial Step for Accurate Overall Survival Estimates
Addressing Postprogression Therapy Confounding in Phase III Oncology Trials: A Crucial Step for Accurate Overall Survival Estimates
Authors: Alexander D Sherry, Pavlos Msaouel, Timothy A Lin, Joseph Abi Jaoude, Ramez Kouzy, Esther J Beck, Avital M Miller, Adina H Passy, Gabrielle S Kupferman, Eugene J Koay, Clifton David Fuller, Charles R Thomas, Zachary R McCaw, Ethan B Ludmir
Introduction:
Overall survival (OS) is the gold standard endpoint in phase III oncology trials, as it directly measures the impact of treatment on patient survival. However, the estimation of the treatment effect on OS can be confounded by the use of post-progression therapies (PPTs) received by patients after the initial disease progression. This cross-sectional study aimed to determine the prevalence of OS analyses accounting for PPT effects in phase III oncology trials.
Methods Overview:
- Cross-sectional analysis of phase III, randomized, controlled oncology trials reported in ClinicalTrials.gov from 2004-2020
- Inclusion criteria: (1) interventional anticancer therapy, (2) two-arm superiority design, (3) reported mature overall survival (OS) findings
- Primary objective: Evaluate the frequency of trials attempting to statistically account for post-progression therapy (PPT) confounding in OS analyses
- Approaches considered: Censoring at PPT initiation, inverse probability of censoring weighting, rank-preserving structural failure, time models
- Recorded whether PPT adjustments were prespecified or post hoc
- Conducted sensitivity analyses for trials evaluating neoadjuvant, definitive, adjuvant, maintenance, first-line metastatic, or first-line hematological settings, as well as trials powered for OS
What We Learned:
The study included 334 phase III trials with 265,310 patients. PPTs were reported in 47% of the trials, but only 12% (41 trials) attempted to account for PPT confounding in the OS analysis. PPT adjustments were often prespecified (56%) and were more likely in trials with cross-over designs (OR 5.04, 95% CI 2.42-10.38) and trials with discordant surrogate-OS findings (OR 2.26, 95% CI 1.16-4.38). In key subgroups, such as trials evaluating locoregional/first-line therapy (8%) and trials powered for OS (11%), PPT analyses were infrequent.
Key Highlights:
- OS analyses accounting for PPT effects were rare, occurring in only 12% of phase III oncology trials.
- Trials with cross-over designs and discordant surrogate-OS findings were more likely to attempt PPT adjustments.
- The most common approach for addressing PPT confounding was simple censoring at the start of PPT, which is generally not recommended due to the risk of selection bias.
- Sensitivity analyses focusing on trials evaluating definitive/first-line therapy and trials powered for OS showed similar low rates of PPT adjustments.
Key Takeaway Messages:
- The effects of post-progression therapies (PPTs) on estimates of the treatment effect on overall survival (OS) may be under-evaluated in phase III oncology studies.
- Several statistical approaches, such as rank-preserving structural failure time models, inverse probability of censoring weighting, and Bayesian methods, can be considered to address PPT confounding, but no single approach has a universal advantage.
- Sequentially multiple randomized assignment trials (SMART) that randomize patients at the time of progression to determine PPT allocation may help control for systematic PPT confounding, particularly in the era of increasing PPT effectiveness.
- Regulatory bodies and oncologists to consider the confounding effects of PPTs when interpreting trial results.
- Future phase III trials to incorporate pre-specified methods for addressing PPT confounding from the outset of trial design.
Summary by Amalya Sargsyan, MD
About OncoDaily
OncoDaily was founded in 2023. It is a US-based oncology media platform, which features the latest news, insights, and patient stories from the world of oncology. Within a short period of time it became one of the leading oncology media platforms globally.
OncoDaily gathers content from various sources, including social media posts from renowned oncologists from all over the world, news from oncology societies and cancer centers, patient and survivor stories, and career-related information for professionals.
The mission of OncoDaily is to empower patients, survivors, and professionals with the knowledge and inspiration they need to fight cancer. The motto of OncoDaily is “Cancer doesn’t take a day off – neither do we.”