IPM Take
This is not just an immunotherapy escalation story. It is a patient-selection story. If dual checkpoint blockade is most powerful in immune-positive tumours, then the future cannot be “treat everyone harder.” It has to be smarter selection, better toxicity prediction and a pathway that knows when escalation is justified.
Executive Summary
Data from the I-SPY2 platform trial, published in JAMA Oncology and highlighted in oncology coverage on 18 August, evaluated cemiplimab plus fianlimab with paclitaxel-based neoadjuvant chemotherapy in high-risk ERBB2-negative early breast cancer. The study included 76 evaluable patients in the intervention arm and 350 in the control arm. Pathologic complete response was 44% with dual checkpoint blockade plus chemotherapy versus 21% in controls across ERBB2-negative disease. In triple-negative breast cancer, pathologic complete response was 53% versus 29%; in hormone receptor-positive / ERBB2-negative disease, it was 36% versus 14%. Among immune-positive tumours, response differences were especially large, including 83% versus 28% in triple-negative breast cancer and 91% versus 28% in HR-positive / ERBB2-negative disease. Adrenal insufficiency occurred in 21% of treated patients, including grade 3 or 4 events in 11%.
Why it matters
- Patients / advocates: Higher response rates matter, but immune toxicity can be long-lasting and life-changing.
- Clinicians: The signal supports further study of dual checkpoint blockade, not indiscriminate escalation.
- Diagnostics / pathology: Immune-positive signatures may become critical if they can identify who benefits most.
- Payers / HTA bodies: Cost and toxicity must be judged against confirmed long-term benefit, not pathologic response alone.
The old oncology reflex is back: the combination looks stronger, so the temptation is to give it to more people.
Breast cancer needs a better reflex.
The I-SPY2 data suggest that adding cemiplimab and fianlimab, targeting PD-1 and LAG-3, can increase pathologic complete response in high-risk ERBB2-negative early breast cancer. The effect was particularly striking in immune-positive tumours, where response rates rose sharply in both triple-negative and hormone receptor-positive disease.
That is exciting. It is also dangerous if read carelessly.
Pathologic complete response is important, especially in high-risk early breast cancer. But it is not the whole patient story. Dual checkpoint blockade can bring immune toxicity that does not vanish when chemotherapy ends. The adrenal insufficiency signal is not a minor inconvenience. It is a warning that escalation has a biological price.
So the right question is not whether dual checkpoint therapy can make tumours disappear more often before surgery. The right question is which patients need that escalation, which patients can avoid it, and how immune signatures can be used without becoming another unevenly available test.
This is where the IPM angle is sharp. Personalised oncology should not only identify targets for more treatment. It should identify who can safely receive less, who needs more, and who needs a different strategy altogether.
If immune-positive signatures are truly predictive, then diagnostics and pathology move to the centre of the story. If they are not implemented consistently, high-capacity centres will select carefully while weaker systems will either overtreat or undertreat.
Dual checkpoint blockade may have a future in early breast cancer. But the pathway must be built around selection, toxicity and proof, not around the thrill of a bigger response number.
