Recurrence-free survival (A, B, E) and OS (C, D, F) based on Kaplan–Meier estimates for subjects treated with GI-4000 or placebo and analyzed for proteomic signature. Baseline samples from 44 subjects in the GI-4000–treated group and 46 subjects in the placebo-treated group were available to investigate the proteomic signature. Samples from GI-4000–treated subjects were further subdivided into training (n = 23) and test (n = 21) sets and all samples not used in the classifier training set were designated BDX-001+ or BDX-001−. Circles, censored data. (A) Recurrence-free survival for GI-4000–treated subjects with (n = 10, black line)/without (n = 11, gray line) BDX-001 signature. (B) Recurrence-free survival for placebo-treated subjects with (n = 18, black line)/without (n = 28, gray line) BDX-001 signature. (C) OS for GI-4000–treated subjects with (n = 10, black line)/without (n = 11, gray line) BDX-001 signature. (D) OS for placebo-treated subjects with (n = 18, black line)/without (n = 28, gray line) BDX-001 signature. (E) Recurrence-free survival for BDX-001+ GI-4000 (n = 10, black line) and placebo-treated subjects (n = 18, gray line). (F) OS for BDX-001+ GI-4000 (n = 10, black line) and placebo-treated subjects (n = 18, gray line).