ROCKET: Phase II Randomized, Active-controlled, Multicenter Trial to Assess the Safety and Efficacy of RRx-001 + Irinotecan vs. Single-agent Regorafenib in Third/Fourth Line Colorectal Cancer

Abstract

Introduction

RRx-001 is a novel cysteine-targeted alkylating agent that releases nitric oxide (NO). The primary biological activities of this hybrid molecule include macrophage repolarizing and vascular normalization. The purpose of this clinical trial (ROCKET) (NCT02096354) was to compare the safety and efficacy of the combination therapy RRx-001 + irinotecan vs. regorafenib in third/fourth line colorectal cancer that previously received treatment with irinotecan.

Patients and Methods

A total of 34 patients were randomized (24 to RRx-001 + irinotecan (RxI) and 10 to single-agent regorafenib (RegI)) and were the basis for the intention-to-treat analysis (ITT, comprising all 34 patients). RRx-001 treatment was administered as an up-to-2-month “primer” followed by irinotecan for patients randomized to the RRx-001 arm (24). The efficacy and safety data are presented for the 34 patients in the (ITT) efficacy analysis. Therapy consisted of intravenous administration of RRx-001 at 4 mg once weekly for up to 2 months, at which point RRx-001 was discontinued, followed by intravenous infusion of irinotecan at 180 mg/m² on day 1 in a 21-day cycle vs. 160 mg oral regorafenib daily for 3/4 weeks followed at progression, if applicable, by irinotecan 180 mg/m² on day 1 in a 21-day cycle. There were 3 patients (3/24 = 12.5%) with prior single agent irinotecan on the RRx-001 randomized arm and 2 (2/10 = 20%) on the regorafenib randomized arm. Numerous patients had irinotecan combination therapies prior to randomized treatment. There were 15 patients on RRx-001 arm that received irinotecan post-RRx-001 in the randomized trial. There were 5 PRs on RRx-001 plus irinotecan leading to an overall response of 20.8% (5/24). There were 37.5% (9/24) of RRx-001 randomized patients with KRAS mutant type while 60% (6/10) regorafenib randomized patients were of KRAS type mutant. There were only 4 patients with available QOL and Edmonton Symptom Assessment System, an insufficient sample size to allow for any meaningful analysis.

Results

Median patient follow-up was approximately 14.5 months (SD 4.5 months). Median overall survival was 8.6 months for RxI and 4.7 months for RegI. Median progression free survival was 6.1 months for RxI vs. 1.7 months for RegI (a statistically significant result, 2-sided log-rank test, P = .0030). The toxicity profile of RxI was substantially improved compared with RegI.

Conclusion

The results of this trial demonstrate improved efficacy of RxI compared with RegI in patients with metastatic colorectal cancer after previous treatment with irinotecan, and late-stage clinical development in this indication is planned on the strength of the observed “signal” accompanied by a sufficient safety profile.