Patient preferences for pancreatic cancer treatment (PERSEUS): a multicenter discrete choice experiment.

[BACKGROUND] Pancreatic cancer has an aggressive nature, and treatment severely impacts patients' quality of life. There is limited understanding how patients weigh treatment benefits against side effects, which hampers the development of patient-centered care and shared decision-making programs.

[METHODS] Two discrete-choice surveys were designed: one comprising pancreatic cancer patients with (borderline) resectable disease (early-stage disease), and one including patients with non-resectable or metastatic disease (late-stage disease). Relevant criteria for describing treatments were identified by literature review and validated through patient and expert interviews. Selected criteria were likelihood of adverse events causing hospitalization, impact on daily functioning, gastrointestinal symptoms, life expectancy and frequency of hospital visits. Interim analysis was executed after 109 inclusions, optimizing the choice task combinations. Patients were recruited from a local center and a nationwide questionnaire project.

[RESULTS] Overall, 428 surveys were sent out and 53% of the participants answered at least one choice task. This included 165 participants with early-stage disease and 62 participants with late-stage disease. Most participants had treatment experience before completing the survey. For both disease stages, participants had a significant preference for the treatment options instead of receiving best supportive care only, although there was significant heterogeneity for this preference among the participants. Life expectancy was the most important treatment characteristic of the pre-selected criteria.

[CONCLUSIONS] Pancreatic cancer patients, both with early-stage and late-stage disease, choose for anti-cancer treatment over best supportive care and value life expectancy as the most important treatment attribute, although significant differences exist between patients.