Longitudinal changes in prognostic awareness (PA) among early phase cancer clinical trial (EPCT) participants.

Abstract

441 Background: EPCT participants may face significant prognostic uncertainty given the nature of the trial and their disease. Understanding how PA evolves during EPCT participation remains understudied and likely has key implications for patient-centered care. Methods: We prospectively enrolled EPCT participants from 04/2021-01/2023. Patients completed baseline (at enrollment) and monthly surveys during trial participation to assess for PA (Prognosis and Treatment Perceptions Questionnaire [PTPQ]; PA dichotomized as “cure my cancer” [inaccurate] vs all other responses [accurate]), symptom burden (Edmonton Symptom Assessment System [ESAS]), quality of life (QOL; Functional Assessment of Cancer Therapy-General [FACT-G]), and hope (Herth Hope Index [HHI]). We used ANOVA to compare mean patient-reported outcome (PRO) scores across PA trajectory groups (i.e., remaining accurate, inaccurate, or changing), Cox regression to explore associations of PA with clinical outcomes (hospitalizations, time on trial [ToT], overall survival [OS]), and chi-square testing to assess palliative care (PC) referrals across PA groups before/during the trial. Results: Among 195 enrolled patients, 192 completed baseline surveys (98% response, median age=63.6 [range: 32-89], 59% female, most common cancer types: gastrointestinal [33%] and breast [20%]). Of the 192 patients who completed a baseline survey, 150 (78%) completed monthly PTPQ surveys. At baseline, 61% (n=91) of patients had accurate PA that remained accurate (A-A), 12% (n=18) had inaccurate PA that remained inaccurate (I-I), 14% (n=21) had inaccurate PA that changed to accurate (I-A), and 13% (n=20) had accurate PA that changed to inaccurate (A-I). We found no significant difference in age, sex, or cancer type across PA groups. Baseline ESAS mean scores and proportion of patients who received 3+ lines of prior therapy significantly differed by PA group (ESAS means: A-A=22.1, I-I=11.7, I-A=19.8, A-I=22.5; p =.033), (3+ lines of prior therapy: A-A=51%, I-I=72%, I-A=19%, A-I=25%; p =.004). Mean FACT-G scores were 73.9 in A-A, 80.1 in I-I, 76.7 in I-A, and 71.3 in A-I groups ( p =.180). Mean HHI scores were 38.9 in A-A, 41.3 in I-I, 39.2 in I-A, and 39.2 in A-I groups ( p =.329). Compared to A-A, the I-A group had longer time to first hospitalization (HR=0.53, p =.048), ToT (HR=0.53, p =.011), and OS (HR=0.48, p =.015); the A-I group had longer ToT (HR=0.46, p =.002). PC referrals occurred in 34% of A-A, 17% of I-I, 30% of A-I, and 29% of I-A groups ( p =.538). Conclusions: We assessed longitudinal PA changes among EPCT participants and identified four PA trajectory patterns (A-A, I-I, I-A, and A-I). Groups with change in PA had longer time to first hospitalization, ToT, and OS. Findings underscore the dynamic nature of PA in EPCT participants and support the need for patient-centered interventions to foster prognostic understanding throughout trial participation.