Nicholas Hornstein, Assistant Professor at Northwell Health, shared a post on LinkedIn:
“COMMIT, now with the full presentation data, deserves a serious re-examination.
Dr. Rocha Lima presents COMMIT
1L dMMR/MSI-H mCRC
Atezolizumab alone vs FOLFOX + bevacizumab + atezolizumab
Key results that stand out:
1.PFS separation is striking
mPFS 24.5 vs 5.3 months
HR 0.439
This looks less like chemotherapy adding benefit and more like underperformance of the PD-L1–only arm, particularly when contrasted with prior PD-1 data from CHECKMATE-8HW and KEYNOTE-177.
2. Risk stratification still matters
Higher-risk patients appeared to do better with the chemo-containing arm, suggesting disease tempo still plays a role even in MSI-H disease.
3. Toxicity is real
Grade 5 events in 9.5% on FOLFOX + bev + atezolizumab is not trivial and sharply limits enthusiasm for universal escalation.
Putting this in context:
CHECKMATE-8HW stopped enrollment early, and at this point 8HW remains the clearest dataset supporting PD-1–based therapy in MSI-H mCRC.
The atezolizumab signal here is surprising. Overall survival looks similar between arms, which may reflect effective post-progression rescue, potentially with dual checkpoint strategies.
Bottom line
This does not argue for chemo-IO for everyone. It reinforces that PD-1 and PD-L1 are not interchangeable, and that PD-1 blockade remains the backbone in MSI-H colorectal cancer.”
More posts featuring Nicholas Hornstein.
