Therapy focus – Thawing conditions for PD-1s in colorectal cancer

Development of immune checkpoint inhibitors for colorectal cancer has been held back by the fact this disease is immunologically “cold” except in a small subgroup of patients known as those with microsatellite instability (MSI).

Although this subgroup is of a sufficient size to justify the development, the field is surely looking for approaches that target a broader range of patients with immuno-oncology combinations, especially as colorectal cancer (CRC) is one of the three most commonly diagnosed solid tumours. While Merck & Co and Bristol-Myers Squibb are focusing on MSI, Roche is taking a combinatorial approach to a broader patient population.

The lack of sensitivity of most CRC patients to immunotherapy is puzzling, given that compared with many other cancer types it has a high mutational frequency, which normally correlates with response to these drugs. Indeed, in terms of its mutational load CRC ranks just behind melanoma and lung cancer, the two indications where immune checkpoint inhibitors have been most successful.

Anti-PD1 agents have also done well in cancers with lower mutation frequencies than CRC, notably kidney and head and neck cancers, so clearly other mechanisms – presumably in the tumour microenvironment – must be involved and will have to be overcome in CRC. 

Microsatellite instability target 

For the time being, Merck & Co and Bristol-Myers Squibb are developing their anti-PD-1 agents for the MSI-high subgroup. Patients with MSI have an exceptionally high mutation burden, but account for 15% of early-stage metastatic CRC and only 4% of later, stage IV disease. 

Merck is more advanced, with two registration studies under way with Keytruda as montherapy: Keynote-164 in previously treated and Keynote-177 in treatment-naive patients. Both recruit only patients who are MSI-high or whose tumours are deficient for DNA mismatch repair (dMMR), from which MSI is a consequence (Asco – The quest for immuno-oncology biomarkers, May 29, 2015).  

Some indication of the activity of Keytruda in the dMMR subgroup emerged at the recent Asco conference in an earlier phase II study. This showed a 57% overall response rate in dMMR CRC, contrasting with a complete absence of responses observed in MMR-proficient CRC tumours (0/25). Merck also reported that it achieved a 62% ORR in the first 13 MSI-high CRC patients in the Keynote-016 study, which is set to render results later this year.  

Meanwhile, Bristol-Myers is focused on a dual PD-1/CTLA-4 blockade strategy in MSI-high patients. It presented data at Asco from CheckMate-142, which tests Opdivo alone and in combination with Yervoy, showing an ORR of 25.5% for monotherapy (70 patients) and 33.3% for the combination (30). Six-month PFS rates were 45.9% for monotherapy and 66.6% for the combination. The company has, however, yet to initiate a registration study in CRC.  

Roche combo

The Roche approach is intriguing, because the group is alone in not selecting patients based on dMMR or MSI; its strategy is to combine Tecentriq with the Mek inhibitor Cotellic.

It presented preliminary results at Asco from a phase Ib trial that showed an ORR of 17%, including four confirmed partial responses, in heavily pretreated CRC patients. Its hypothesis is that Cotellic is sensitising CRC tumours to Tecentriq by increasing MHC I expression on tumour cells and promoting intratumoral CD8+ T-cell accumulation.

Roche noted that all responders in its study had Kras mutations, and the responses did not correlate with MSI status. It has just launched a 360-patient phase III trial, Cotezo, to evaluate Tecentriq alone or in combination with Cotellic against Bayer’s Stivarga in third-line or later CRC.

AstraZeneca and Pfizer/Merck KGaA have yet to disclose any plans to develop their respective PD-1/PD-L1 blockers durvalumab or avelumab in CRC, although there is an investigator-sponsored study under way with the former.  

Later-stage studies in CRC with PD-1/PD-L1 inhibitors 
Project Study Enrolment Setting Design  Trial ID Data
Keytruda  Keynote-177 270 1L, MSI-H or dMMR Vs standard therapy NCT02563002 Aug 2019
Keytruda  Keynote-164 120 Tx-experienced, MSI-H or dMMR Monotherapy (no control) NCT02460198 May 2017
Opdivo  CheckMate-142 96 MSI-H Opdivo +/- Yervoy (no control) NCT02060188 Sep 2016
Tecentriq  - 360 >3L  +/- Cotellic vs Stivarga NCT02788279 Apr 2019

Given the size of the potential market for immuno-oncology agents in CRC it is surprising that there are not more combinations in late-stage development, and it could be that advances can still come from targeted agents in this field.

One new possible triple combination recently emerged when Array announced plans at Asco to study encorafenib and binimetinib combined with Erbitux in Braf-mutant CRC patients (Asco 2016 – Array to test Braf/Mek combo in colorectal cancer, June 5, 2016). It would seem that in this particular indication there is still much to play for.  

To contact the writer of this story email Robin Davison at or follow @RobinDavison2 on Twitter

Share This Article