PD-L1 testing is a lab test performed on tumor tissue to measure how much of a specific protein, called Programmed Death-Ligand 1, is present on or around cancer cells. The result helps oncologists determine whether a patient is likely to respond to a type of cancer treatment called immunotherapy. It’s now a standard part of the diagnostic workup for several cancers, including non-small cell lung cancer, bladder cancer, and triple-negative breast cancer.
Why PD-L1 Matters in Cancer
Your immune system has built-in brakes that prevent it from attacking your own healthy tissue. One of those brakes involves a protein called PD-L1 on cell surfaces. When PD-L1 connects with a receptor called PD-1 on immune cells, it sends a “stand down” signal. That’s useful for protecting normal tissue, but cancer cells exploit this system. Many tumors produce PD-L1 on their surface as a kind of disguise, effectively telling immune cells to leave them alone.
This is sometimes called “immune escape.” The cancer cell tricks the immune system into treating it like healthy tissue, allowing the tumor to grow unchecked. Immunotherapy drugs called checkpoint inhibitors work by blocking that PD-1/PD-L1 connection, stripping away the tumor’s disguise so the immune system can recognize and attack it. PD-L1 testing tells your oncologist how much of that protein the tumor is producing, which helps predict whether these drugs will work for you.
How the Test Is Done
PD-L1 testing is performed on a tissue sample, usually from a biopsy or surgical specimen. The tissue is preserved in a wax block (called a formalin-fixed, paraffin-embedded block), sliced into thin sections, and placed on a glass slide. A pathologist then uses a technique called immunohistochemistry, which applies specially designed antibodies that bind to the PD-L1 protein. When the antibodies attach, they produce a visible stain under the microscope, allowing the pathologist to see which cells are producing PD-L1 and how many.
There are specific quality standards the tissue must meet. The sample needs to contain at least 100 viable tumor cells, and the tissue block should ideally be less than 3 years old. Older specimens or those with too few tumor cells can underestimate how much PD-L1 is actually present, potentially leading to a falsely low result. In some cases, a pathologist may recommend a new biopsy if the available tissue doesn’t meet these requirements.
Results typically come back quickly. Despite a dramatic increase in testing volume over recent years, most patients receive their PD-L1 result within 2 to 4 days, with an average turnaround time of about 3 days.
How Results Are Scored
PD-L1 results aren’t reported as a simple positive or negative. Instead, pathologists use scoring systems that quantify how much PD-L1 is present. The two most common are:
- Tumor Proportion Score (TPS): The percentage of viable tumor cells that stain positive for PD-L1. If 50 out of 100 tumor cells show staining, the TPS is 50%.
- Combined Positive Score (CPS): This counts PD-L1-positive tumor cells plus PD-L1-positive immune cells (like certain white blood cells) in and around the tumor, divided by the total number of viable tumor cells, multiplied by 100. CPS captures a broader picture of the immune environment.
Which scoring system your pathologist uses depends on the type of cancer and which immunotherapy drug is being considered. TPS is most commonly used in lung cancer, while CPS is frequently used for gastric, cervical, and head and neck cancers.
What the Numbers Mean for Treatment
Different cancers and different drugs use different cutoffs, but a few thresholds come up repeatedly. In non-small cell lung cancer, a TPS of 50% or higher qualifies a patient for pembrolizumab as a first-line treatment (assuming no certain genetic mutations are present). A TPS of 1% or higher may still open the door to immunotherapy in combination with chemotherapy.
For gastric and gastroesophageal cancers, a CPS of 5 or higher is increasingly recognized as the threshold that optimizes the balance between benefit and risk when adding immunotherapy to chemotherapy. At a CPS of 10 or higher, the benefit appears even more pronounced, with one large trial showing a 35% reduction in the risk of death in that subgroup. Patients whose tumors score below a CPS of 1 generally don’t see a meaningful survival benefit from adding checkpoint inhibitors.
These cutoffs aren’t universal rules. They vary by cancer type, the specific drug being considered, and evolving clinical evidence. Your oncologist interprets the score in the context of your particular diagnosis.
Which Cancers Require PD-L1 Testing
PD-L1 testing is currently required or recommended across a growing list of cancer types. The FDA has approved companion diagnostic tests for non-small cell lung cancer, urothelial (bladder) cancer, head and neck squamous cell cancer, gastric and gastroesophageal junction cancers, esophageal squamous cell cancer, triple-negative breast cancer, and cervical cancer. For some of these, testing is mandatory before prescribing specific immunotherapy drugs. For others, testing is available but not strictly required for treatment decisions.
Not All PD-L1 Tests Are the Same
One complicating factor is that there are multiple FDA-approved PD-L1 tests, each developed with a different antibody and paired with a specific immunotherapy drug. The 22C3 assay is used as a companion diagnostic for pembrolizumab across multiple tumor types. The SP142 assay is paired with atezolizumab. The 28-8 assay is used for the combination of nivolumab and ipilimumab. And the SP263 assay serves as a companion diagnostic for atezolizumab in certain settings.
These assays don’t always produce identical results on the same tissue. Each antibody has slightly different binding characteristics and staining patterns. This means that a PD-L1 score generated with one assay may not be directly interchangeable with a score from another. In practice, the assay your lab uses will typically match the drug your oncologist is considering.
Why a Single Test Can Be Misleading
PD-L1 expression is not uniform throughout a tumor. Research comparing different sites in the same patient found that PD-L1 status between a primary tumor and a metastatic site matched only about 61% of the time. That means in nearly 4 out of 10 patients, a biopsy from one location would give a different PD-L1 reading than a biopsy from another.
PD-L1 levels also shift over time. In patients who received chemotherapy before retesting, PD-L1 status between their pre-treatment and post-treatment tumors matched only 57 to 63% of the time. Whether this reflects the direct effects of chemotherapy or the tumor’s natural evolution isn’t entirely clear, but the practical implication is the same: a PD-L1 result represents a snapshot, not a fixed characteristic of the cancer.
One study concluded that at least five biopsy samples may be necessary to accurately capture a tumor’s PD-L1 status in advanced gastric cancer, since the protein’s expression can vary even within different regions of the same tumor. This kind of variability is one reason oncologists sometimes make treatment decisions based on multiple factors beyond PD-L1 alone, including tumor genetics, overall health, and the specific cancer type.