Patients' and physicians' risk-benefit trade-off preferences for metastatic colorectal cancer treatments

Abstract

Background: Preferences for treatment-related benefits and risks associated with metastatic colorectal cancer (mCRC) therapies may vary between patients and physicians. Methods: A literature review and clinician interviews were carried out to develop risk-benefit profiles. PFS and probabilistic adverse outcomes were described in an online discrete-choice experiment completed by patients and physicians. Participants assessed a series of 10 choices between pairs of hypothetical medication profiles . Each profile included attributes within a pre-determined range: PFS (12 months, 8 months, 6 months), severe papulopustular rash (PR) (0%, 5%, 10%, 25%), serious hemorrhage (0%, 2%, 10%, 35%), cardiopulmonary arrest (0%, 2%, 10%, 20%), and gastrointestinal perforations (0%, 2%, 10%, 20%). Choice questions were based on an experimental design with known statistical properties. Random-parameter choice models produced preference weights indicating the strength of trade-off. These weights were used to calculate the maximum acceptable risk of different adverse events associated with various mCRC therapies. Results: A total of 127 patients and 150 physicians completed the discrete-choice survey. The mean maximum level of treatment-related risk patients were willing to accept for a 4-month increase in PFS from 8 to 12 months was 16.7% (0.0%-82.9%) for PR, 13.8% (4.6%-23.2%) for gastrointestinal perforation, 10.3% (3.3%-17.1%) for serious hemorrhage and 5.1% (1.7%-8.7%) for cardiopulmonary arrest. For the same PFS improvement, physicians were willing to accept a risk of severe PR that exceeded 25% and more than 20% risk for gastrointestinal perforation, which were the maximum levels shown in the survey for each of these treatment-related adverse events. Physicians were also willing to accept 18.9% (12.3%-24.5%) risk for serious hemorrhage and 7.0% (3.0%-10.8%) for cardiopulmonary arrest for the same 4-month PFS improvement. Conclusions: The benefit-risk tradeoff data obtained in this study show potential risk-tolerance differences between patients and oncologists, highlighting the importance of understanding preferences from the patient’s perspective when making treatment decisions.