Faculty Perspectives: Rationale for PD-L1 Expression as a Biomarker in Immuno-Oncology | Part 2 of a 4-Part Series

← Back to Issue

Ensuring Consistent and Reliable Biomarker Testing

Harris S. Goodman, MD

Biomarkers and the Cancer Care Team: Critical Partners in Personalized MedicineIn the main article of this publication, the current state of programmed-cell death-1 and programmed-cell death ligand-1 (PD-L1) testing, evasion by tumors of the immune system, and the study and use of immune checkpoint inhibitors across multiple tumor types are expertly summarized. Because this area of oncology is undergoing massive investigation, rapid progression, and change, it is imperative that physicians (especially oncologists), oncology nurses, oncology nurse navigators, and pharmacists are aware of these changes and respond appropriately. To do otherwise could subject their patients to unnecessary and costly testing, inappropriately prescribed treatment (and its associated costs and side effects), and possibly worse outcomes.

As stated in the article, tissue-based PD-L1 assays were developed in parallel with specific therapeutic agents as companion biomarkers, and this has led to considerable variability in PD-L1 testing. Although it is not necessary for the treating oncologist or nurse to know which staining platform was used to perform the test, it is essential that they know which factors affect the test result and thus the decision to treat. Preanalytical variables, such as the quality of the specimen, the choice of fixative, and the rapidity of appropriate fixation and processing, may significantly affect immunohistochemical test results, and are often overlooked.1,2 For example, cytologic specimens—such as malignant pleural and peritoneal fluids—may be fixed with an alcohol-based fixative, as opposed to tissue samples that are routinely fixed with 10% neutral-buffered formalin. Most immunoperoxidase stains are validated only on formalin-fixed, paraffin-embedded tissues and, therefore, results are only valid on such tissues. One must be acutely aware of which immuno­peroxidase antibody clone is being used, because each clone is directly associated with a specific treatment (eg, anti–PD-L1 clone 28-8 and nivolumab, antibody clone 22C3 and pembroliz­umab, etc), and different clones have different cutoffs to determine eligibility for, and expected response to, treatment.

In our laboratory, we require the oncologist to inform us of which immune checkpoint inhibitor is being considered for treatment before we will perform a PD-L1 immunohistochemistry test. This ensures the appropriate test is being performed and also prevents waste of scarce financial resources. Attempting to apply the results of one PD-L1 assay associated with a specific treatment to a different unassociated treatment is not recommended, at least until more rigorous studies have been performed evaluating interchangeability of assays with real-world samples. Similarly, knowing which tumor types are currently appropriate for PD-L1 testing is necessary, and although the list of these tumors continues to grow,3 testing is not indicated irrespective of tumor type.

Interobserver variability is a difficult issue to address and may be a significant problem when the PD-L1 positivity threshold is at 1%. Choosing a laboratory that has passed repeated rigorous accreditation requirements (there are several accrediting organizations in the United States that perform these inspections) may help ensure consistent, high-quality testing and reliable results. This is rarely, if ever, a question asked by clinicians; it is most definitely the responsibility of the laboratory leadership and laboratory medical director, regardless of whether they are performing the test in-house or sending the tissue specimen out to a specific reference laboratory for their PD-L1 testing.4

PD-L1 testing, and other prognostic and predictive markers, are somewhat unique in the immunoperoxidase testing realm because a single test determines eligibility for a very specific treatment. Most immunoperoxidase stains are performed as a panel, often for identification of cells of origin of a tumor, and thus failure of one stain in a panel does not typically lead to misidentification of a tumor and inappropriate treatment. Similarly, estrogen and progesterone immunohistochemistry markers often stain in tandem. HER2 by immunoperoxidase often (but not always) reflects, in an inverse fashion, breast cancer differentiation, with well-differentiated tumors infrequently yielding an HER2-positive result. However, PD-L1 immunoperoxidase tests are typically performed in isolation, with results having a significant impact on patient care. It is ultimately up to the cancer care team to decide whether the test result should be used to guide therapy. Presumably, this decision is made before initiating testing, but questionable results should always be investigated.

Malignant tumors change with both time and treatment; it is likely that PD-L1 results for a particular tumor will also change. Hopefully, future studies will better elucidate if this is the case. The cancer care team will have to decide, with the input of the patient, whether a subsequent biopsy is indicated upon disease progression and whether repeat prognostic and predictive marker testing, including PD-L1, should be performed. In addition, as alluded to in the main article of this publication, incorporation of other biomarkers—such as tumor mutation burden and microsatellite instability status—is already playing a significant role in the choice of treatment and thus the patient’s course.

Members of the cancer care team must maintain clear communication with each other regarding patient status, biomarker test results, pharmacologic therapy, tumor response to treatment, and adverse events to optimize desired outcomes, minimize setbacks, and utilize financial resources effectively and responsibly.


  1. Fitzgibbons PL, Bradley LA, Fatheree LA, et al. Principles of analytic validation of immunohistochemical assays: guideline from the College of American Pathologists Pathology and Laboratory Quality Center. Arch Pathol Lab Med. 2014;138:1432-1443.
  2. Gown AM. Diagnostic immunohistochemistry: what can go wrong and how to prevent it. Arch Pathol Lab Med. 2016;140:893-898.
  3. Patel SP, Kurzrock R. PD-L1 expression as a predictive biomarker in cancer immunotherapy. Mol Cancer Ther. 2015;14:847-856.
  4. Department of Health and Human Services, Centers for Medicare and Medicaid Services. Clinical laboratory improvement amendments of 1988; final rule. Fed Register. 2003(Jan 24):[42CFR493.1252(a)].
Uncategorized - August 23, 2018

The Evolving Role of PD-L1 Testing in Patients with Cancer

Clearly, immunotherapy has come of age in multiple tumor types, as described in the main article in this publication. The question now is how to develop more personalized immunotherapy and use biomarkers to identify patients most likely to benefit from this therapy. To benefit from immunotherapy, a patient’s tumor should [ Read More ]

Uncategorized - August 23, 2018

Rationale for PD-L1 Expression as a Biomarker in Immuno-Oncology

Therapy with immune checkpoint inhibitors (ICIs) heralds a new era in the management of many cancers. In particular, immunotherapy directed toward cytotoxic T-lymphocyte–associated antigen 4 and the programmed-cell death-1 (PD-1) receptor and its ligand, PD-L1, has demonstrated remarkable and durable responses in several types of cancer, including melanoma, non–small-cell lung [ Read More ]