Finding Harmony in PD-L1 Testing


Marlon Rebelatto

Marianne Ratcliffe

Jill Walker

Members of the durvalumab diagnostic team recently travelled to the European Society for Medical Oncology (ESMO) 2016 Congress in Copenhagen, Denmark, to present our industry-leading research on programmed cell death ligand-1 (PD-L1) testing.1

A team of scientists from AstraZeneca and MedImmune worked together to deliver this study, where 500 tumour samples from patients with head and neck cancer were evaluated using three validated PD-L1 diagnostic assays to assess the concordance between tests.1

The team had presented similar data earlier in the year at the American Association for Cancer Research (AACR) with data from a similar, large scale study in NSCLC.

Why does concordance matter? There are a number of PD-1 and PD-L1-directed antibodies emerging as therapies in multiple settings such as NSCLC and head and neck cancer, and each therapy has its own diagnostic test.1-3 Understanding the comparability of these different assays is critically important not only for determining patient suitability for monotherapy with checkpoint inhibitors, but also for identifying patients who may require combination immune-oncology (IO) treatments to help increase likelihood of response. The results of our study showed that the three leading assays agree very well across the whole assay range in both NSCLC and head and neck cancer, which builds optimism that they may be able to be used interchangeably in the future. Our data will enable conversations with regulatory authorities in support of a target of all patients having access to high quality, robust PD-L1 testing.1,4

PD-L1 Expression is Not Binary: Until now, PD-L1 expression has generally been conceptualized as a binary variable: patients are classified as either PD-L1 positive or negative. 5 Depending on the potential therapy, however, both the test and the appropriate cut-off point for classifying a patient as PD-L1 positive may differ. 5,1 PD-L1 expression is not like ALK or EGFR testing where a mutation is either present or absent; it is a continuous variable, and PD-L1 expression should be viewed along a continuum.5,6

Moving forward, due to the variability in PD-L1 classification and patient response it is highly likely that multiple cut-offs will become clinically relevant when making decisions around treatment. 5,1 Depending on the proposed therapy, the optimal level of PD-L1 expression necessary to increase the chances of a response to the drug will differ. Clinical data on the patient’s tumour and the therapy will help determine the appropriate cut-off.

Looking Ahead: At AstraZeneca and MedImmune, we are continuing our leading research in PD-L1 testing concordance, working with trained pathologists to evaluate data from multiple validated assays.1 This research supplements the work of the Blueprint initiative, a cross-industry working group developed to review the issue of harmonization between PD-L1 diagnostic tools that also presented its first data set at the AACR meeting in April. 7,8

We know that PD-L1 status is informative for monotherapy in lung cancer, and we believe it will become increasingly important as we evaluate immuno-oncology combination therapies for a wide variety of patients.1 While additional research is needed, we are excited about AstraZeneca/MedImmune’s leadership role in biomarker testing, and ultimately, bringing IO therapies to more patients who may benefit.



  1. Ratcliff M et al. A comparative study of PDL1 diagnostic assays in squamous cell carcinoma of the head and neck (SCCHN). ESMO 2016 Poster. To be presented October 2016.
  2. Pardoll D et al. The blockade of immune checkpoints in cancer immunotherapy. Nature Reviews Cancer; 12: 252- 264. April 2012. doi: 10.1038/nrc3239
  3. Postow M et al. Immune Checkpoint Blockage in Cancer Therapy. J Clin Oncol 33. Online First 20 January 2015. DOI: 10.1200/JCO.2014.59.4358
  4. Ratcliff M et al. A comparative study of PDL1 diagnostic assays and the classification of patients as PDL1 positive and PDL1 negative. AACR 2016 Poster abstract. Presented April 2016.
  5. Patel S and R Kurzrock. PD-L1 Expression as a Predictive Biomarker in Cancer Immunotherapy. Mol Cancer Ther; 14: 847-856. Published Online First 18 February 2015. doi:10.1158/1535-7163.MCT-14-0983
  6. Awad M and P Hammerman. Durable Responses With PD-1 Inhibition in Lung and Kidney Cancer and the Ongoing Search for Predictive Biomarkers. Journal of Clinical Oncology Vol 33, 2015. Published online on 27 April 2015. doi: 10.1200/JCO.2015.61.4172
  7. Blueprint Consortium. The Blueprint Project: Comparing PD-L1 IHC diagnostics for Immune Checkpoint Inhibitors. AACR 2016 Regulatory Science and Policy Session Oral. Presented April 2016.
  8. Blueprint Consortium. The Blueprint Project: Comparing PD-L1 IHC Diagnostics for Immune Checkpoint Inhibitors. AACR 2016 Regulatory and Science and Policy Session Abstract. Presented April 2016.

Page Atlas ID: 1013134.011
Date of preparation: 06 October 2016
Date of expiry: 06 October 2017