Aakash Desai, Associate Director, Phase 1 and Precision Oncology Program at UAB O’Neal Comprehensive Cancer Center, shared on LinkedIn:
“A recurring tension in first-line metastatic NSCLC: we know adding tremelimumab to durvalumab + chemo improves overall survival (POSEIDON), but the absolute benefit is modest, and not everyone needs the second checkpoint.
So the real question has always been: who?
TRIDENT, out now in AACR Journals Clinical Cancer Research, takes a run at it. Ferdinandos Skoulidis and colleagues trained machine-learning models on clinical, genomic, and radiomic data from POSEIDON to find the patients who gain the most OS from adding tremelimumab.
What stood out to me:
- Clinical + genomic data carried the signal. Radiomics were in the mix but didn’t do the lifting, which is a useful reality check on multimodal hype.
- In the top-ranked 50% of non-squamous tumors, OS HR was 0.56 (95% CI 0.33–0.97).
- The benefit signature includes KRAS and STK11 mutations. STK11 is the subset we typically flag as immunotherapy-resistant, so a CTLA-4 rescue signal there may make sense.
Caveat: This is a post-hoc analysis of a single, industry-sponsored trial (AstraZeneca), and that confidence interval sits right against 1.0 (95% CI 0.33–0.97). It’s a hypothesis to test prospectively. Good to see this line of work moving in the direction of integrating ML and AI, and it’ll be models like this, weighing clinical, genomic, and eventually imaging features together to tell us not just whether to give IO, but which combination, and to whom.”
Title: Utilizing machine learning to identify multimodal signatures for patients who would benefit from the addition of tremelimumab to durvalumab and chemotherapy (TRIDENT)
Authors: Ferdinandos Skoulidis, Salma K. Jabbour, Edward B. Garon, Puneeth Iyengar, Giorgio Scagliotti, Loïc Ferrer, Guillaume Etchepare, Olivier Gallinato, Jérôme Faure, Paul Bernard, Thierry Colin, Philippe Menu, Yian Lin, Ling Cai, Ammar Ahmed Chaudhry, Amanda Remorino, Ross Stewart, Luisa Luciani-Silverman, Katy Miller, David Dellamonica, Jolyon Faria, Yiduo Zhang
Other articles featuring Aakash Desai on OncoDaily.