Background: There are no validated blood-based biomarkers to monitor efficacy in pts with advanced melanoma. Lactate dehydrogenase (LDH) is an established prognostic factor; however, it is not normally used to inform treatment decisions. ctDNA at baseline (BL) is associated with a poor prognosis in pts treated with BRAF inhibitors, but no studies have examined the association between serial changes in ctDNA and survival after BRAF and/or MEK inhibitor therapy. Methods: We measured BRAF V600E/K ctDNA at BL and wk 4 in plasma samples from a pooled population of pts with unresectable or metastatic melanoma treated with D or D+T in the phase 3 COMBI-d trial (NCT01584648). We used mutation-specific droplet digital PCR assays; ctDNA results were categorized as positive/negative (pos/neg) using an analytically validated detection threshold of 0.25 copies/mL. Progression-free (PFS) and overall survival (OS) were analyzed in all pts and by BL LDH level (> or < upper limit of normal). Results: BL ctDNA was detectable in 320/345 pts (92.7%) and was not associated with survival. Nearly all pts had a considerable drop in ctDNA after 4 wks of therapy; 201 pts had paired samples (BL and wk 4) and detectable ctDNA at BL. In 80/201 pts (40%) whose ctDNA changed from pos at BL to neg at wk 4, PFS and OS were prolonged vs in 121/201 (60%) whose ctDNA remained pos (median PFS, 12.9 [95% CI, 9.2-20.2] mo vs 7.1 [5.5-8.9] mo; HR, 0.55 [0.39-0.76]; P = .0003; median OS, 28.2 [20.5-48.8] mo vs 14.6 [11.8-18.7] mo; HR, 0.56 [0.40-0.79]; P = .0007). Undetectable ctDNA at wk 4 was associated with prolonged PFS and OS, especially in pts with high BL LDH (Table). Conclusions: Particularly in pts with high LDH, on-treatment ctDNA monitoring may be helpful for early identification of pts likely to benefit from D or D+T. All ctDNA Samples at Wk 4. Clinical trial information: NCT01584648