Disease	Colorectal cancer
Stage of Disease/Treatment	Metastatic/advanced
Prior Therapy	2 prior regimens
Type of Study	Phase I, 3+3 dose‐limiting toxicity (phase Ia) and expansion (phase Ib)
Primary Endpoint	Safety and tolerability
Secondary Endpoints	Pharmacokinetics, efficacy, biomarker associations (exploratory)
Additional Details of Endpoints or Study Design
This was a phase Ia/b, multicenter (three sites across the U.S. and France), nonrandomized, open‐label study. The study comprised a phase Ia DLT observation part (for one treatment cycle) and a phase Ib expansion part. The DLT observation part followed a 3+3 design, in which if one of three patients experienced a DLT, three additional patients were to be enrolled. Patients who completed the observation part without a DLT continued study treatment until a criterion for discontinuation was met. The expansion part started when up to one of the six patients treated developed a DLT. In the expansion period, 15 additional patients were to be enrolled at the recommended dose for the expansion cohort identified in the DLT observation part. Thus, the treatment regimen during the expansion part was the same as during the DLT observation part.
The main inclusion criteria were: age ≥18 years; advanced or mCRC; measurable disease at the time of study enrollment; had received prior second‐line treatment with oxaliplatin and/or irinotecan (no prior immune checkpoint inhibitors had been administered and patients with RAS wild‐type CRC must also have received prior treatment with an epidermal growth factor receptor monoclonal antibody); Eastern Cooperative Oncology Group (ECOG) performance status of 0 to 1; and adequate organ function. The main exclusion criteria were significant gastrointestinal bleeding within 3 months and significant venous thromboembolic events or any arterial thromboembolic events within 6 months prior to enrollment; uncontrolled hypertension; treatment with chronic nonsteroidal anti‐inflammatory drug or antiplatelet therapy at the time of enrollment; or other serious uncontrolled medical disorders.
The primary endpoints for safety and tolerability were DLTs and TEAEs. DLTs were defined as grade 4 hematologic toxicity persisting >5 days; grade ≥3 febrile neutropenia; grade 4 thrombocytopenia (unless recovered in 24 hours and in the absence of bleeding) or grade 3 thrombocytopenia complicated with grade ≥2 bleeding; grade 3 nonhematologic toxicity occurring despite maximal supportive medical management; or any other clinically significant toxicity deemed by the investigator and the sponsor's clinical research physician to be dose limiting (such as grade 2 seizures or severe tremors). TEAEs were coded per the Medical Dictionary for Regulatory Activities version 21.1. For phase Ib, the final analysis occurred approximately 1 year after the last patient received his or her first dose of study treatment. Predetermined adverse events of special interest (AESIs) included infusion‐related reactions, hypertension, proteinuria, thromboembolic events, bleeding or hemorrhagic events, gastrointestinal perforation, reversible posterior leukoencephalopathy syndrome, congestive heart failure, fistula formation, surgery and impaired wound healing, and liver failure or liver injury.
Pharmacokinetic endpoints were the minimum concentrations of ramucirumab (serum) and merestinib (plasma), which were analyzed using a validated enzyme‐linked immunosorbent assay and a validated protein precipitation method, respectively.
Preliminary efficacy endpoints included overall response rate (proportion of patients who achieved a complete response [CR] or partial response [PR] as their best overall response) and PFS. OS was also evaluated. Scans for restaging were performed every 6 weeks (±7 days) for the first 6 months after enrollment and every 9 weeks (±7 days) thereafter until radiographic disease progression, death, or study completion, whichever occurred first.
Exploratory endpoints included PFS and OS by biomarker status. Plasma collected at baseline was evaluated for circulating cell‐free tumor DNA using the Archer® Reveal ctDNA™ 28‐gene next generation sequencing assay (ArcherDX, Inc., Boulder, CO), which evaluates 28 genes including HRAS (exons 2, 3), KRAS (exons 2, 3, 4), NRAS (exons 2, 3), and TP53 (all exons). Samples were classified as HRAS, KRAS, NRAS, or TP53 mutation positive if the analysis identified genetic variants that are confirmed somatic/germline mutations in publications annotated in COSMIC [1] or predicted somatic/germline mutations via the FATHMM algorithm [2]; samples were classified as negative if analysis identified no variants or only variants of unknown significance. VEGF‐A, VEGF‐D, and soluble VEGFR3 (sVEGFR3) concentrations were evaluated as described previously [3] in plasma samples collected at baseline.
Safety analyses included all patients who received one or more doses of ramucirumab and/or merestinib. Efficacy analyses included all patients who received any quantity of study treatment (intent‐to‐treat analysis). Pharmacokinetic analyses included all patients who received one or more doses of ramucirumab and merestinib and had baseline and one or more postbaseline evaluable pharmacokinetic samples. PFS and OS were estimated by Kaplan‐Meier methodology. Cox regression analysis was performed to evaluate the association between response outcomes and biomarkers (treated as binary outcomes; mutation positive vs. negative for genetic markers and high vs. low for VEGF‐A, VEGF‐D, and sVEGFR3 [dichotomized at the median concentration for each marker]). Multiplicity adjustment was applied, adjusted for all biomarkers. The data cutoff date was November 5, 2018. Statistical analyses were performed using SAS software (version 9.4; SAS Institute, Cary, NC).
Investigator's Analysis	Drug tolerable, hints of efficacy