Abstract 6257
Background
Primary analysis of COMBI-AD (NCT01682083) showed significant improvement in RFS with adjuvant D + T vs placebo (Pbo; HR, 0.47; P < .001). We present an RFS analysis with extended follow-up (FU), a cure-rate model, and biomarker analysis.
Methods
COMBI-AD is a randomized, phase 3 trial evaluating 12 mo of adjuvant D + T vs Pbo in pts with resected BRAF V600–mutant stage III melanoma. A Weibull mixture cure-rate model was used to estimate the proportion of pts who will not relapse. Mutational landscape and gene expression signatures (GES) were examined in baseline tissue samples by sequencing 570 genes and gene expression profiling (GEP) with a NanoString® panel.
Results
Median FU was 44 mo (D + T) and 42 mo (Pbo); 177/438 pts (41%) in the D + T arm and 254/432 pts (59%) in the Pbo arm had relapsed/died. Median RFS was not reached (NR; 95% CI, 46.9 mo-NR) with D + T vs 16.6 mo (95% CI, 12.7-22.1 mo) with Pbo (HR, 0.49 [95% CI, 0.40-0.59]). Estimated cure rate was 54% (95% CI, 49%-59%) with D + T vs 37% (95% CI, 32%-42%) with Pbo. DNA sequencing, GEP, and paired DNA + RNA results were available for 368, 507, and 301 pts, respectively. MAPK pathway gene alterations did not correlate with outcome. Immune GES (eg, interferon [IFN]-γ signature) were strongly prognostic in both arms. High tumor mutational burden (TMB) added positive prognostic value to IFN-γ signature in the Pbo arm (high IFN-γ and high TMB associated with longer RFS), whereas in the D + T arm, IFN-γ gene signature identified pts with longer RFS independently of TMB status. Exploratory analysis of RFS in D + T vs Pbo arms in all TMB/IFN-γ subgroups suggested that low TMB or high TMB/high IFN-γ may be associated with greater RFS benefit than high TMB/low IFN-γ.
Conclusions
The updated RFS and cure-rate model confirm the ongoing benefit of adjuvant D + T. MAPK gene alterations previously associated with targeted therapy resistance were not associated with outcome in the adjuvant setting. TMB and immune GES identified pts at higher risk of relapse in the Pbo arm. The predictive value of these markers with targeted therapy or checkpoint inhibitors warrants further validation in a prospective study.
Clinical trial identification
NCT01682083
Editorial Acknowledgement
Medical writing assistance was provided by Michael Demars, PhD of ArticulateScience, LLC and funded by Novartis Pharmaceuticals Corporation