Evaluating Patient Preferences for Clinical Trial Endpoints in Early Stage Cancer: A Discrete Choice Experiment in Canada
Author(s)
Alexis T. Mickle, MSc1, Frances Simbulan, MSc2, Bianca Li, BSc2, Kristoph Klein-Panneton, MSc2, Stephanie Snow, MD3, Conor Morrison, MSc1, Karissa M. Johnston, PhD1.
1Broadstreet HEOR, Vancouver, BC, Canada, 2AstraZeneca Canada, Mississauga, ON, Canada, 3Queen Elizabeth II Health Sciences Centre, Halifax, NS, Canada.
1Broadstreet HEOR, Vancouver, BC, Canada, 2AstraZeneca Canada, Mississauga, ON, Canada, 3Queen Elizabeth II Health Sciences Centre, Halifax, NS, Canada.
OBJECTIVES: Although patient preferences for non-overall survival (OS) measures in advanced cancers have been previously assessed, less is known for early-stage disease. This study aimed to estimate preferences and trade-offs for safety and non-OS endpoints for people with early-stage, curable cancer.
METHODS: A moderator-lead discrete choice experiment (DCE) was conducted in April 2025 among Canadian adults with early-stage cancers. Participants completed ten DCE choice sets, each comparing standard-of-care (SoC) and a new treatment. Attributes and levels were informed by qualitative interviews, literature review, and clinical expert input. Attributes included five-year OS, two-year disease-free progression (DFP), pathological complete response (pCR), and short- and long-term side-effects. The new treatment had unknown OS by definition. The SoC treatment remained static in the choices shown to each respondent, while all attributes except OS varied across choice sets for the new treatment with each participant randomly assigned one SoC OS value. Data were analyzed using logistic regression and presented as odds ratios (ORs) with 95% confidence intervals (CIs).
RESULTS: Among 103 adults treated for early-stage breast (n=32), lung (n=31) or gastrointestinal (n=40) cancer, median (Q1, Q3) age was 59 (43, 75) years; 46.6% were female. Participants were 3.49 times more likely to select a new treatment for every 25% decrease in SoC OS (OR, 3.49; CI, 2.82-4.31; p<0.01); 55% more likely to select a new treatment for every 25% increase in DFP (1.55; 1.26-1.91; p<0.01); 85% less likely to select a treatment with severe short-term side-effects (ref: mild/no side-effects; 0.15; 0.10-0.23; p<0.01); and 10% less likely to select a treatment for every 25% increase in long-term side-effects (0.90; 0.81-1.00; p<0.05).
CONCLUSIONS: When survival data were available (SoC treatments), OS played an important role in treatment-making decisions. For new treatments where OS was not yet known, respondents placed high importance on measures of disease progression and safety.
METHODS: A moderator-lead discrete choice experiment (DCE) was conducted in April 2025 among Canadian adults with early-stage cancers. Participants completed ten DCE choice sets, each comparing standard-of-care (SoC) and a new treatment. Attributes and levels were informed by qualitative interviews, literature review, and clinical expert input. Attributes included five-year OS, two-year disease-free progression (DFP), pathological complete response (pCR), and short- and long-term side-effects. The new treatment had unknown OS by definition. The SoC treatment remained static in the choices shown to each respondent, while all attributes except OS varied across choice sets for the new treatment with each participant randomly assigned one SoC OS value. Data were analyzed using logistic regression and presented as odds ratios (ORs) with 95% confidence intervals (CIs).
RESULTS: Among 103 adults treated for early-stage breast (n=32), lung (n=31) or gastrointestinal (n=40) cancer, median (Q1, Q3) age was 59 (43, 75) years; 46.6% were female. Participants were 3.49 times more likely to select a new treatment for every 25% decrease in SoC OS (OR, 3.49; CI, 2.82-4.31; p<0.01); 55% more likely to select a new treatment for every 25% increase in DFP (1.55; 1.26-1.91; p<0.01); 85% less likely to select a treatment with severe short-term side-effects (ref: mild/no side-effects; 0.15; 0.10-0.23; p<0.01); and 10% less likely to select a treatment for every 25% increase in long-term side-effects (0.90; 0.81-1.00; p<0.05).
CONCLUSIONS: When survival data were available (SoC treatments), OS played an important role in treatment-making decisions. For new treatments where OS was not yet known, respondents placed high importance on measures of disease progression and safety.
Conference/Value in Health Info
2025-11, ISPOR Europe 2025, Glasgow, Scotland
Value in Health, Volume 28, Issue S2
Code
PCR81
Topic
Patient-Centered Research
Topic Subcategory
Patient-reported Outcomes & Quality of Life Outcomes
Disease
Oncology, Pediatrics