Do you offer pembrolizumab monotherapy or carbo-pemetrexed-pembrolizumab to a patient with Stage IV large cell NSCLC with PD-L1 of 80% and KRAS+STK11 mutation?

This is a good question. My opinion is first that the KRAS is irrelevant and only studied because of the mouse model. The entire effect is from LKB1, independent of KRAS. Second, that this has not been tested in any of the randomized trials looking at IO vs. IO/chemo vs. chemo, so its predictive ability has not been demonstrated. I would try single agent pembro until further data are available.

I would approach this patient as usual with a discussion of risks (ie toxicities) and benefits (ie efficacy) of each course of therapy. Focusing just on the questions of montherapy with pembrolizumab vs the combination, I would focus on the "toxicities leading to death" in the respective studies (Table 2 in Keynote-024 and Keynote-189). For monotherapy, this rate was 0.6% (1/154) vs 2.0% for chemotherapy (3/150). In the Keynote-189 trial on the other hand, the rate was 6.7% for chemo-pembrolizum...