Case 3: PD-L1 Testing Used in Triple-Negative Breast Cancer

Hope Rugo, MD, FASCO: Another area that concerns us when we’re selecting patients is what tests we’re using for PD-L1. You mentioned that you use the Ventana PD-L1 (SP142) assay, Adam. Evita, is that what you’re using? And do you do that at your own institution, or is it a send-out test?

Evita Sadimin, MD: For triple-negative breast cancer, we do use the SP142. We don’t have Ventana in-house, so that one is a send-out test. For the purpose of PD-L1 testing, we only have the 22C3 test, so that is the one we offer in-house. However, when we have the order for SP142 test for eligible triple-negative breast cancer, then we send it out.

Hope Rugo, MD, FASCO: There is some discussion about the fact that it’s all in the eyes of the observer, so being able to detect 1% positivity might vary a lot. What is your thought on that?

Evita Sadimin, MD: It is definitely subjective, and I’m not going to lie about that. But there is a methodical approach to scoring these specimens, and different clones also have different ways of scoring. For example, for 22C3 we do the combined positive score [CPS], where you have to take into account the tumor cells and the immune cells within the tumor and outside the tumor cells. But for the SP142 test, you’re only scoring the immune cells within the tumor area.

Hope Rugo, MD, FASCO: I don’t think any of us realized there would be such a big variation between tests. I think seeing the KEYNOTE-355 data when it’s presented will be very interesting to try and understand what the correlation is between a CPS score of 10 or greater and SP142 of 1%. It may be that these are sort of crossover interactive scoring systems.

Ian Krop, MD, PhD: Up until 6 months ago, it wasn’t quite clear. You had these 2 tests with clearly different sensitivities. I think the data you presented at ESMO [European Society for Medical Oncology]—where you looked at the IMpassion130 data broken down by the different antibodies—was quite compelling in that it really does appear that the added number of patients who are positive by the pembrolizumab antibody doesn’t seem to contribute any additional sensitivity. It’s really the ones who are positive by both antibodies. That is where the real benefit of the checkpoint inhibitor is. That makes me much more comfortable right now in terms of the testing.

Adam Brufsky, MD, PhD: The other thing that was really fascinating about your presentation at ESMO was the fact that PD-L1 testing is different depending on the tissue type or where the metastasis is from. For some reason, liver metastases have less. I think that is something to consider. If I get a liver metastasis and it’s negative, I’m really thinking, “Is it truly negative?” That’s a really interesting piece of data. I don’t know if it’s true or not, but it’s something that a lot of us have looked at.

Hope Rugo, MD, FASCO: I think that is interesting.

Transcript edited for clarity.