ASCO-GI 2019—Durvalumab-tremelimumab extends OS in refractory CRC

  • Melissa Pandika
  • Univadis
Access to the full content of this site is available only to registered healthcare professionals. Access to the full content of this site is available only to registered healthcare professionals.


  • Durvalumab combined with tremelimumab extended OS by 2.5 months relative to best supportive care (BSC) alone in a phase 2 study of patients with advanced refractory colorectal cancer (rCRC).

Why this matters

  • Until now, immune checkpoint inhibitors had not shown activity in patients with rCRC, except in those with defective mismatch repair (dMMR) tumors.

Study design

  • Canadian Cancer Clinical Trials Group (CCCTG) CO.26 phase 2 trial of 179 patients with advanced CRC refractory to all available therapy, randomly assigned 2:1 to:
    • Durvalumab (1500 mg every 28 days until disease progression) + tremelimumab (75 mg every 28 days for first 4 cycles only) and BSC.
    • BSC alone
  • No patients with known dMMR tumors were enrolled.
  • Primary endpoint was OS, with a 2-sided P-value

Key results

  • Median follow-up was 15.2 months.
  • Median OS was 6.6 months for durvalumab + tremelimumab and BSC vs 4.1 months for BSC (P=.07).
  • Disease control rate was 22.7% for durvalumab + tremelimumab and BSC vs 6.6% for BSC (P=.006).
  • Yet there was no effect on PFS (median, 1.8 vs 1.9 months; P=.97).
  • Significantly less deterioration on EORTC QLQ-C30 physical function for durvalumab + tremelimumab and BSC than BSC alone at 16 weeks.
  • Adverse events were more frequent for durvalumab + tremelimumab and BSC.


  • Nonblinded, non-placebo-controlled.
  • Designed with alpha of 0.1, allowing 10% false positive risk.
  • Extra follow-up in treatment arm.
  • Discordant efficacy results.
  • Confirmation of dMMR status is ongoing. 

Expert comment

  • “This is a positive trial statistically. However, I think there's a lot of caveats.... I think fundamentally before this data is incorporated into any clinical practice, we'd clearly need confirmation of these findings to have confidence in them,” said Michael J. Overman, MD, professor of gastrointestinal medical oncology, The University of Texas MD Anderson Cancer Center, Houston, Texas, who was not involved in the trial.