Improved decision making from clinical trial evidence in an era of personalised oncology

The program of research reported in this thesis addresses challenges in clinical trial design arising in the era of personalised oncology. Specifically: the value of progression free survival (PFS) as a surrogate for overall survival (OS) in first line trials of advanced ovarian cancer (Ch 2); the use of a control group and subgroup analyses in a randomised phase 2 trial of regorafenib in advanced oesophago-gastric cancer (Ch 3); the value of central review of imaging when used as key end point (Ch 4); and, the ability of a nomogram to improve prognostication in advanced colorectal cancer trials (Ch 5). \nMETHODS \nMeta-analysis of published data from randomised phase 3 trials (Ch 2); primary and secondary analyses of randomised phase 2 clinical trial with primary end point of PFS (Ch 3); evaluation of central review for key end points in a randomised phase 2 trial (Ch 4); development of nomograms for PFS and OS from a database of randomised trials in advanced colorectal cancer (Ch 5). \nFINDINGS \nCh2: PFS may not be a good surrogate for benefits on OS in first line ovarian cancer trials where participants are later treated with multiple lines of active salvage therapies. However, PFS may be useful for understanding heterogeneity of treatment effects across subgroups in this setting. \nCh 3: PFS was a useful measure of the activity of Regorafenib, a targeted therapy, as second- or third-line treatment for advanced oseophago-gastric carcinoma. PFS was also a useful indicator of the homogeneity of the treatment effect of regorafenib in most subgroups, but raised the hypothesis of a larger treatment effect in residents of South Koreans than of Australia, New Zealand, or Canada. The control group was helpful because a priori estimates from historical controls proved inaccurate. \nCh 4: treatment effects based on investigator-assessed PFS were accurate, unbiased, and robust in INTEGRATE (1). Central review of PFS in INTEGRATE (1) was feasible and corroborated the trial conclusions. Effective blinding with placebo, and a large treatment effect, meant that substantial error would have been needed for central review to change conclusions. \nCh 5: A practical risk assessment tool was developed, providing prognostic information useful for risk stratification in future clinical trials, and for guiding personalised decision-making in routine practice, for patients with advanced/metastatic colorectal cancer commencing first line systemic therapy. \nCONCLUSIONS \nThis research has demonstrated considerations pertinent for determining: the usefulness of PFS as an indicator of activity, and to determine effects across subgroups; the validity and need for central review in a placebo-controlled trial; baseline risk assessment for better trials and clinical decision making.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.