There has been much frustration on many fronts at the number of trials that do not see a relationship between PD-L1 expression and response. Some do, but many don’t. This has lead to quite a few investigators suggesting that the IHC assay may not be as useful as originally hoped, for predicting response to checkpoint blockade or selecting patients for therapy.
While we often do see a trend for more responders with higher levels of expression, the main issue is that PD-L1-negative patients can also see some responses, albeit at a lower rate.
There are many factors that can affect the measurement:
- Fresh vs. archival tissue
- Heterogeneity within the tumour
- Tumour cells (TC) vs. immune cells (IC)
- Different antibodies used for each assay
- The dynamic nature of the tumour microenvironment – does timing of the biopsy matter?
- Human error – a pathologist has to eyeball the IHC readouts and decide the level of staining intensity
And so on. These are just a few examples of the factors that can potentially affect the results, making it quite a challenging test to undertake. There is also time – does the level of expression vary temporally depending on which prior therapies are administered?
It would be easy to be disheartened by this, but fear not!
There were some impressive new data presented at ASCO that were not only intriguing, but also show us a way forward on how a multi-factorial approach could be used in different tumour types. By this I mean we might end up with different tests used in conjunction for several different cancers in order to a) predict responders and non-responders and b) better select patients for appropriate regimens or clinical trials.
It’s not going to be as easy as one size (or test) fits all. Sometimes a more more sophisticated approach will be needed. New data at ASCO gave us hints on what’s to come in this direction.
To learn more about these new developments, subscribers can log-in below or if interested in an individual or corporate license, click on the blue box below…