Association Between Circulating Hepatocyte Growth Factor (hgf) and Clinical Response Across Five Studies with the Braf Inhibitor Dabrafenib or Mek Inhibitor Trametinib

e20019 Background: To investigate the previously reported association between baseline HGF and clinical response in patients (pts) treated with RAF inhibitors, HGF levels were assessed in pt blood from the dabrafenib (D) and trametinib (T) single agent and combination (D+T) trials BRF113710, BRF112680, BRF113683, MEK114267, and BRF113220. Methods: Circulating levels of HGF were assessed in baseline pt plasma using Aushon BioSystems SearchLight ELISA. Within and between arm Cox and logistic regression models were utilized to assess the association between HGF levels and progression-free survival (PFS) and response rate (RR). The statistically optimal cut-point was determined (Simon 1995). The relationships between HGF levels and baseline covariates (SLD: sum of longest diameters; LDH: lactate dehydrogenase) were assessed using Spearman’s correlation, Wilcoxon’s test (ECOG), and covariate adjusted Cox models (when data permit). Analyses were conducted separately for each study. Results: Pts with ECOG 1 or 2 had higher levels of HGF than pts with ECOG 0. HGF was weakly positively correlated with LDH and SLD. HGF was not associated with RR. Higher levels of HGF were associated with shorter PFS in pts treated with D in 3 of 4 studies, but were not associated with PFS with T alone or in combination with D (Table). Optimal cut points were in the 60th-70th HGF percentile. HGF was not an independent predictor of PFS. Results for overall survival are being analyzed. Conclusions: High levels of HGF in circulation were predictive of shorter PFS in response to D monotherapy. The impact of HGF on PFS was not observed when pts received T or D+T, and HGF was not an independent predictor when adjusted for covariates. PFS PFS covariate adjusted Study N Hazard ratio (HR) P value (P) HR P D BRF113710 68 1.53 0.0068 1.35 0.1549 BRF112680 20 2.43 0.0110 NA BRF113683 179 1.42 0.0002 0.97 0.9066 BRF113220 21 1.31 0.5082 NA T MEK114267 186 1.09 0.3512 1.10 0.3862 Control arms BRF113683 (DTIC) 54 1.03 0.8897 0.97 0.9066 MEK114267 (CHEMO) 89 1.56 0.0006 1.37 0.0027 D+T BRF113220 (1 mg) 26 1.29 0.4867 NA (2mg) 24 0.73 0.4835 NA NA = Insufficient data for covariate models.