With PD-L1 Under Fire, Attention Turns to Better Immunotherapy Biomarkers

Paolo Bossi, MD; Floriana Morgillo, MD, PhD; Stefan Zimmermann, MD; Markus Joerger, MD, PhD, ClinPharm


October 10, 2016

Editor's Note:
One of the most interesting and hotly debated issues at ESMO 2016 has been the future of programmed death-ligand 1 (PD-L1) as a predictive biomarker for immunotherapy.

Earlier studies have already shown that PD-L1 is far from perfect. In light of this, several expert contributors to Medscape's ESMO 2016 Live Blog have turned their attention to novel alternatives presented in scientific and educational sessions at ESMO 2016.

For a comprehensive review of the most important ESMO 2016 sessions across all tumor types, visit Medscape's ESMO 2016 Live Blog.

Problems With PD-L1 as a Predictive Biomarker

Markus Joerger, MD, PhD, ClinPharm: There is a big issue with PD-L1 immunoexpression as a predictive biomarker for patients receiving PD-(L)1–targeting monoclonals, one that has been demonstrated in the Blueprint Proposal presented at the American Association for Cancer Research earlier this year: Different assays cannot easily be compared with each other. The reason is the different monoclonal antibodies used for PD-L1 staining, the differing thresholds put forth, and the different target cells (ie, tumor cells vs immune-infiltrating cells).

In a session at ESMO titled "How to Select Patients for Immunotherapy: The Role of Pathology," Mogens Vyberg, MD, of Aalborg University Hospital in Denmark, outlined the complexity of defining an optimal threshold for defining PD-L1–positive vs PD-L1–negative, and therefore characterized PD-L1 immunostaining as an "imprecise biomarker." So far, the only situation in which an immune checkpoint inhibitor is conditionally approved for patients being PD-L1–positive is pembrolizumab as second-line treatment in patients with advanced non–small-cell lung cancer (NSCLC).

Dr Vyberg showed results of the NordiQC immunostaining harmonization study that identified important potential pitfalls for successful immunostaining—not limited to PD-L1 staining—with the most important ones being insufficiently calibrated antibody dilutions and erroneous epitope retrieval. Overall, efforts are ongoing to cross-validate PD-L1 immunostaining on tumor cells by the different monoclonal antibodies.

Solange Peters, MD, PhD, from Lausanne University Hospital in Switzerland, covered the important issue of the controversial predictive value of PD-L1 immunostaining in advanced NSCLC. This topic is of particular interest if we consider the relatively new approval of nivolumab and pembrolizumab in patients with advanced NSCLC, the high number of patients with NSCLC, and the high costs of the immune checkpoint inhibitors. Dr Peters asked the important question of whether PD-L1 is the current biomarker to use in this group of patients. Again, the complexity of different drugs (nivolumab, pembrolizumab, avelumab, atezolizumab, druvalumab), different target cells (tumor cells and immune infiltrates), and different thresholds was raised, meaning that cross-validation has become increasingly difficult. According to the Blueprint Proposal, concordance of the different assays is only about 50%.

Another controversy is the fact that some clinical trials, such as POPLAR (atezolizumab vs docetaxel second-line treatment) or CheckMate 057 (nivolumab vs docetaxel second-line treatment), showed that patients with PD-L1 staining < 1% did not get a benefit from treatment with the anti–PD-(L)1 monoclonal antibody. Still, there are reasons to treat even PD-L1–negative patients with second-line checkpoint inhibitors, because of superior tolerability or the fact that patients may still respond to treatment despite being PD-L1–negative.

For NSCLC frontline treatment, the first clinical trial data have just been presented at this year's ESMO annual meeting, and these data suggest that selecting patients for PD-L1 positivity seems to be important. The KEYNOTE-024 study showed that pembrolizumab first-line monotherapy is superior to standard chemotherapy in terms of progression-free and overall survival in patients with high (> 50%) PD-L1 expression.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: