Clinical reference

Biomarker Directory
Back to directory

PD-L1

Established

PD-L1 IHC is one of three FDA-approved predictive biomarkers for immune checkpoint inhibitor response.

Also known as PD-L1 CD274 B7-H1
In plain language

PD-L1 is a protein some tumors use to hide from the immune system, and testing for it helps predict whether a patient will benefit from an immunotherapy drug that blocks this hiding mechanism. A real complication: different immunotherapy drugs each require their own specific PD-L1 test, and even pathologists reviewing the identical sample don't always agree on the score.

Biomarker & context

PD-L1 (programmed death-ligand 1, gene CD274) is a cell-surface protein that suppresses T-cell activity when bound to its receptor PD-1; tumor expression of PD-L1 is one of three currently FDA-approved predictive biomarkers for immune checkpoint inhibitor response, alongside TMB and MSI-H/dMMR.

What it indicates

PD-L1 expression predicts response to PD-1/PD-L1 checkpoint inhibitors (pembrolizumab, nivolumab, atezolizumab, and others) across NSCLC, gastric, cervical, head and neck, urothelial, and triple-negative breast cancer, among others; higher expression generally correlates with greater likelihood of benefit, though response still occurs at low or even negative expression in some settings.

Diagnosis & clinical testing

Measured via immunohistochemistry using tumor proportion score (TPS, percentage of tumor cells staining positive) or combined positive score (CPS, includes tumor-associated lymphocytes and macrophages); critically, each checkpoint inhibitor has its own FDA-approved companion diagnostic antibody clone (22C3 for pembrolizumab, SP142 for atezolizumab, 28-8 for nivolumab, SP263 for durvalumab), and these clones are not interchangeable despite testing the same protein.

Guiding treatment

For first-line pembrolizumab monotherapy in advanced NSCLC, a TPS ≥50% is required; lower thresholds (TPS ≥1%) support use in other settings or combination regimens. A real, well-documented limitation: even FDA-cleared assays show only moderate inter-rater agreement at certain cutoffs (kappa as low as 0.32 for CPS ≥1% in one comparison study), meaning two pathologists can reasonably disagree on the same slide.

Clinical significance

Each checkpoint drug has its own non-interchangeable companion diagnostic antibody clone and scoring system (TPS vs. CPS).

References
  1. ARUP Consult. (2024). PD-L1 Testing: Choose the Right Test.
  2. Mayo Clinic Labs. (2024). Biomarker and companion diagnostic testing.
  3. Yarchoan, M., et al. (2025). The complexities of PD-L1 expression as an indicator of immunotherapy outcomes.