Atez-er for cancer imaging.

It’s difficult to reliably predict which cancers will respond to PD1-axis inhibitors. Some contenders for best predictive capabilities include tumor expression of PD-L1 and tumor mutational burden. Now a new trick has been thrown in the ring: a PET scan tagged, not with glucose, but with the anti-PD-L1 agent atezolizumab. This early-phase study imaged 22 patients with untreated advanced bladder, non-small cell lung, or triple-negative breast cancer using zirconium-89-labeled-atezo. Note that serum concentration of atezo for imaging were roughly 100-fold less than therapeutic doses and was found to be quite safe. Interestingly, uptake was high in tumors at all sites but varied substantially within tumors, perhaps explaining why tumor expression of PD-L1 in a single biopsy sample does not perfectly predict response to therapy. Most interestingly, all imaged patients went on to receive atezo monotherapy, and pre-treatment max SUV was significantly correlated with treatment response while traditional tumor expression of PD-L1 per immunohistochemistry was not. TBL: Kudos to Roche for a brilliant new potential market grab. | Bensch, Nat Med 2018