Posted: April 2018
Solange Peters, MD-PhD, PD-MER, runs the Thoracic Malignancies Program in the Department of Oncology at the University of Lausanne, Switzerland. Her main field of interest is new biomarker discovery and validation in preclinical/clinical settings, and she is also involved in developing multimodality trials for locally advanced NSCLC. She has been elected President of the European Society for Medical Oncology (ESMO) for the 2020-2021 term.
Here, Dr. Peters discusses the global importance of the Blueprint Project’s efforts regarding standardization of PD-L1 assays and terminology. She also defines tumor mutation burden (TMB) and shows how it might complement PD-L1 and speculates on the evolving complexity of biomarker identification and use in predicting immune response.
Q: How has the Blueprint Project affected thoracic oncology practice in Europe? Are there similar or duplicative projects in Europe?
A: The Blueprint Project has changed the practice in Europe. There are similar initiatives at national, or even regional, levels in Europe. The initiatives— a German ring trial,1 a French initiative,2 and a few others in Europe and beyond3,4—started more less at the same time as the Blueprint Project, with results being reported around the same time. It is very reassuring that they are all reporting similar results and describing, with accuracy, the same conclusions. The Blueprint Project is important because we don’t usually speak about potential and probable regional/interinstitutional differences and variations in the way immunohistochemistry might be performed and interpreted. I think that the convergence of the regional data in combination with the international reliability of the IASLC Project is extremely strong. All of the initiatives aim to harmonize the determination of PD-L1 positivity. This picture has been confused and confounded by the various pharmaceutical companies that were each using a different way to assess PD-L1, at every step of this assessment methodology. With the publication of the Blueprint results, there is now a stronger effort to try to standardize the way PD-L1 is defined. For example, every university hospital in Switzerland, which is a small country, is using the same platform, the same antibody, and the same quality-assessment processes including round robins to ensure that PD-L1 is being defined exactly the same way across the country.
Q: To what extent will TMB supplement or supplant PD-L1 testing?
A: We don’t have a prospective evaluation of the predictive ability of TMB yet; the only data we have come from convergent retrospective analyses. However, what we have learned from these retrospective data is that TMB will be a very good biomarker in the future and that it is independent from PD-L1. The two biomarkers don’t correlate but might complement each other; they basically describe two different phenomena. When you speak about TMB, you speak about the potential presence of neoantigens, which might elicit an immune response—it’s about how immunogenic a tumor is supposed to be. On the other hand, when you look at PD-L1, it’s related to the end process. It describes how much the immune response has really been taking place in the tumor after the many steps of activation, presentation, cross-priming, and trafficking of T cells. Therefore, you describe two different processes: one focuses on tumor immunogenicity, and the other on effector-phase microenvironment. Both are interesting in trying to predict the immune sensitivity of tumors, and testing for both might be recommended practice in the future. We know from the press release that Checkmate-227 will probably prospectively prove that TMB is predictive.5 I still haven’t seen real data trying to combine the use of both biomarkers (TMB and PD-L1). At the end of the next few years, we might not have just one biomarker, but a kind of signature for predicting the immunogenicity and the potential immune sensitivity of tumors.
Q:What is your take on gene signature analyses as well as other putative predictors of response to immunotherapy agents?
A: Use of immune signatures is closer to use of PD-L1 expression in that you try to look in the tumor cells and microenvironment and see how much the immune interaction with the tumor has given rise to immunogenic reaction. When you look at immune–gene signature— at least the ones that have been prospectively assessed until now—you describe a kind of surrogate of PD-L1 in that the same end process is described. Interestingly, the IMpower150 trial, presented at ESMO 2017, showed that the immune–gene signature was no more predictive than PD-L1 regarding patient selection and therapeutic outcome.6 I’m still not convinced that the immunerelated genes we’ve been using add anything to PD-L1. On the contrary, TMB, immunogenic signatures, and PD-L1 will probably correlate very strongly. For the time being, I think we should look for better and stronger immune– gene signatures or expression profiles.
I think we have to determine how we can better characterize and qualify the interaction of the cancer cells with the microenvironment, which will teach us what we want to examine and how we want to go about this.
Q: Are there any other predictors on the horizon?
A: TMB testing is fundamentally just about calculating the probability of the presence of some strong epitopes in the tumor, suitable for immune system recognition—it’s just a probabilistic approach. There is a huge effort ongoing to identify the actual neoantigens in tumors, and there are a lot of algorithms being developed to try to predict the best epitopes for immunoreactivity. You can also test the tumor and immune system of a patient to see if it recognizes this epitope, but this is complex, and data are disappointing to date. Future research might be focused on identification of new antigens, potentially building on that by expanding the pool of T cells to attack them. We are quite far from there, however, and researching this is going to be tricky, expensive, and time consuming. I think, for the time being, we have already done quite a lot of work to try to harmonize TMB and to learn how to measure it and how to make it a practically approachable option for immuneresponse prediction; I think this is going to be the main area of progress for the next 2 to 3 years.
Q: Do you have any other comments?
A: Complexity continues to increase regarding immunotherapy agents now that combinations are being used. When we start to use new immunotherapy drugs like IDO inhibitors or other immune modifiers (LAG3, OX40, TIM3), for example, we may end up delineating patient populations by their markers. A patient’s signature for PD-1/PD-L1 will be rendered even more complex by the fact that, when combination therapy is used, there is involvement of a second signaling pathway or protein that is targeted by the second drug. ✦
References:
1. Scheel AH, Dietel M, Heukamp LC, et al. Harmonized PD-L1 immunohistochemistry for pulmonary squamous-cell and adenocarcinomas. Mod Pathol. 2016; 29(10):1165-1172.
2. Adam J, Rouquette I, Damotte D, et al. PL04a.04: Multicentric French Harmonization Study for PD-L1 IHC Testing in NSCLC. J Thorac Oncol. 2017;12(1) :S11-S12.
3. Rimm DL, Han G, Taube JM, et al. A Prospective, Multi-institutional, Pathologist-Based Assessment of 4 Immunohistochemistry Assays for PD-L1 Expression in Non-Small Cell Lung Cancer. JAMA Oncol. 2017;3(8):1051-1058.
4. Ratcliff e MJ, Sharpe A, Midha A, et al. Agreement between Programmed Cell Death Ligand-1 Diagnostic Assays across Multiple Protein Expression Cutoffs in Non- Small Cell Lung Cancer. Clin Cancer Res. 2017;23(14):3585-3591.
5. Statement: Bristol-Myers Squibb Validates Predictive Role of Tumor Mutation Burden in Phase 3 CheckMate -227 Program in First-Line Non-Small Cell Lung Cancer. Bristol-Myers Squibb website. news.bms.com/press-release/corporatefinancial-news/bristol-myers-squibb-validates-predictive-role-tumor-mutation-. Accessed February 9, 2018.
6. Reck M. Primary PFS and safety analyses of a randomized phase III study of carboplatin + paclitaxel +/- bevacizumab, with or without atezolizumab in 1L non-squamous metastatic NSCLC (IMpower150) [Abstract LBA1_PR]. Ann Oncol 2017(28Suppl 11).
