Onvansertib in Combination with FOLFIRI and Bevacizumab in Second-Line Treatment of KRAS-Mutant Metastatic Colorectal Cancer: A Phase Ib Clinical Study

Abstract

Purpose: Onvansertib is a highly specific inhibitor of polo-like kinase 1 (PLK1), with demonstrated safety in solid tumors. We evaluated, preclinically and clinically, the potential of onvansertib in combination with chemotherapy as a therapeutic option for KRAS-mutant colorectal cancer.

Patients and methods: Preclinical activity of onvansertib was assessed (i) in vitro in KRAS wild-type and -mutant isogenic colorectal cancer cells and (ii) in vivo, in combination with irinotecan, in a KRAS-mutant xenograft model. Clinically, a phase Ib trial was conducted to investigate onvansertib at doses 12, 15, and 18 mg/m2 (days 1-5 and 14-19 of a 28-day cycle) in combination with FOLFIRI/bevacizumab (days 1 and 15) in patients with KRAS-mutant metastatic colorectal cancer who had prior oxaliplatin exposure. Safety, efficacy, and changes in circulating tumor DNA (ctDNA) were assessed.

Results: In preclinical models, onvansertib displayed superior activity in KRAS-mutant than wild-type isogenic colorectal cancer cells and demonstrated potent antitumor activity in combination with irinotecan in vivo. Eighteen patients enrolled in the phase Ib study. Onvansertib recommended phase II dose was established at 15 mg/m2. Grade 3 and 4 adverse events (AE) represented 15% of all treatment-related AEs, with neutropenia being the most common. Partial responses were observed in 44% of patients, with a median duration of response of 9.5 months. Early ctDNA dynamics were predictive of treatment efficacy.

Conclusions: Onvansertib combined with FOLIFRI/bevacizumab exhibited manageable safety and promising efficacy in second-line treatment of patients with KRAS-mutant metastatic colorectal cancer. Further exploration of this combination therapy is ongoing. See related commentary by Stebbing and Bullock, p. 2005.

Figure 1.

Effect of onvansertib monotherapy and in combination with irinotecan in colorectal cancer preclinical models. A, Isogenic KRAS MUT and KRAS WT DLD1 cells were treated at the indicated concentrations of onvansertib for 72 hours. Bar graph (mean ± SEM, n = 3) represents relative viability to DMSO for each cell line. B, KRAS MUT and KRAS WT DLD1 cells were treated for 24 hours with DMSO or 100 nmol/L onvansertib. Bar graph (mean ±SEM, n = 3) represents the percentage of mitotic cells measured by flow cytometry (positive staining for phospho-H3 Ser28). Data analyzed by two-way ANOVA using Šídák multiple comparisons test. C and D, BALB/c nude mice inoculated with HCT116 cells were treated with vehicle, onvansertib, irinotecan, or the combination for 42 days. C, Mean tumor volumes ± SEM, n = 7–8/group. Tumor volumes were compared at day 42 using a one-way ANOVA with Tukey multiple comparison tests. D, Mean ± SEM body weight changes from baseline. *** and **** indicate P ≤ 0.001 and 0.0001, respectively.

Figure 2.

Clinical efficacy of onvansertib in combination with FOLFIRI/bevacizumab. A, Swimmer plot according to dose level. Each line represents a patient. B, Waterfall plot of maximum percent change in target lesions from baseline. C,KRAS mutation at baseline according to best response. D, Kaplan–Meier curve of PFS. The gray shadow indicates 95% CI. The data are from n = 16 patients evaluable for efficacy. PR = partial response, SD = stable disease, PD = progressive disease.

Figure 3.

Early changes in KRAS MAF are associated with clinical benefit. A, Percentage change in KRAS MAF after one cycle of treatment from baseline according to best response. Fifteen patients were included in the analyses, 7 with PR, and 8 with either SD or PD as best response. Data analyzed by unpaired t test with Welch correction, **, P < 0.01. B, ROC curve for clinical response prediction, showing the sensitivity and specificity of KRAS-mutant ctDNA decrease after one cycle to predict clinical response. 95% CIs are shown in blue. C, Patient with ≥90% decrease in KRAS MAF from baseline were considered molecular responders. Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of the KRAS-mutant ctDNA plasma test to predict clinical response are reported.