Early quality of life (QOL) and symptom analysis from the DREAMseq phase III randomized control trial of combination immunotherapy versus targeted therapy in patients (pts) with BRAF-mutant metastatic melanoma (MM) (ECOG-ACRIN EA6134).

9559 Background: Combinations of either immune checkpoint inhibitors (anti-PD-1/anti-CTLA-4) or BRAF/MEK-targeted therapies have shown significant clinical benefit in pts with BRAFV600 mutant MM. Until recently, little prospective data existed to guide the choice of initial therapy or sequence. Results of the DREAMseq Trial showed that the treatment sequence beginning with nivolumab/ipilimumab (Nivo/Ipi) immunotherapy produced a clinically meaningful 20% improvement in 2-year overall survival (OS) compared to the sequence beginning with dabrafenib/trametinib (Dab/Tram) targeted therapy. The OS and progression-free survival (PFS) curves were biphasic crossing at 10 and 6 months, respectively. Our aim is to characterize QOL trends within and between the initial therapies through 24 weeks (wks). Methods: 265 pts were randomly assigned to Nivo/Ipi for up to 12 wks then Nivo alone (Arm A) or Dab/Tram continuously (Arm B) and at disease progression (PD) received the alternate therapy. QOL was assessed by the PROMIS Profile 29 at baseline, wk 12 (end of cycle (C) 2), and wk 24 (end of C4). Wilcoxon Signed Rank test was used to examine changes over time within treatment arms. OS was estimated by Kaplan-Meier method to compare between pts who stopped treatment for toxicity on Arm A by C2 and who continued on Arm A therapy to C4. A complete case analysis compared QOL domain means for (C2) vs. (C4). Pt-reported adverse events were also collected. Results: Baseline completion rates for the PROMIS-29 for Arm A (n = 108, 81.2%) and Arm B (n = 117, 88.6%) and decreased to 28.6% and 53.8%, respectively at C4. Through C4, the principal reasons for dropout were toxicity (35.2% for Arm A and 11.9% for Arm B) and PD (26.1% for Arm A and 18.6% for Arm B). From Baseline to C2: Arm B reported statistically significant improvements in Pain Interference (-3.45, P = 0.007), Sleep (-2.11, P = 0.014), and Anxiety (-3.74, P < 0.001). By C4, these early differences had dissipated (mean diff. = 0.73 – 1.73, all p = NS). For pts remaining on treatment to C4 (n = 157), a complete case analysis indicates no significant QOL differences between C2 vs C4. Pts stopping for toxicity on Arm A after C2 had similar 2-yr OS to pts who continued on Arm A to C4. QOL at C2 (Arm A: stopping for toxicity vs. on treatment) were meaningful, but underpowered (Physical Health (PH) mean difference = -3.5, p = 0.18). Conclusions: Over the first 12 wks, Dab/Tram is associated with significant improvement in overall function and less disturbance in in sleep, pain, physical function, and PH than Nivo/Ipi as expected by PFS curves and toxicity profiles. These differences dissipate by 24 wks when Arm A therapy has switched to Nivo alone and PFS curves cross. Early QOL and treatment cessation due to Nivo/Ipi toxicity was not associated with differences in 2-yr OS. Clinical trial information: NCT02224781.