A team from Hospital Universitario 12 de Octubre evaluates PD-L1 expression and tumor-infiltrating lymphocyte densities in early-stage NSCLC by comparing manual pathology with Navify Digital Pathology and PathAI algorithms. Their AI-assisted workflow speeds turnaround, improves reproducibility, and identifies more PD-L1–positive cases at clinically relevant cutoffs.
Key points
Navify Digital Pathology SP263 and PathAI AIM-PD-L1-NSCLC algorithms achieve ICC>0.98 for continuous PD-L1 TPS versus manual consensus.
AI tools detect significantly more cases with ≥1% PD-L1 TPS (p=0.00015), affecting immunotherapy eligibility.
PathAI and Navify TIL algorithms show strong correlation (r=0.49) between total H&E TILs and CD8+ cell densities.
Why it matters:
AI-driven pathology scoring promises faster, more reproducible biomarker quantification in NSCLC, enabling better patient selection for immunotherapies.
Q&A
What is PD-L1?
What are tumor-infiltrating lymphocytes?
What is Tumor Proportion Score (TPS)?
How do AI algorithms improve pathology workflows?
Why measure turn-around time (TAT)?
Read full article
Academy
Programmed Death Ligand 1 (PD-L1)
What is PD-L1? PD-L1 is a protein on the surface of some cells, including cancer cells, that interacts with the immune system to regulate attack. When PD-L1 binds to its receptor PD-1 on T cells, it sends a “don’t kill me” signal that suppresses immune activity.
Why is PD-L1 important in lung cancer? In non-small cell lung cancer (NSCLC), levels of PD-L1 on tumor cells guide the use of immune checkpoint inhibitors—drugs that block PD-1/PD-L1 interactions and unleash patient T cells against cancer. Higher PD-L1 expression typically predicts better responses to these immunotherapies.
Tumor-Infiltrating Lymphocytes (TILs)
What are TILs? TILs are immune cells, chiefly T cells, that penetrate the tumor and surrounding stroma. A high density of CD8+ TILs often indicates an active immune response against the tumor.
How are TILs measured? Pathologists examine stained tissue sections under a microscope or use digital scans. Automated algorithms can count CD8+ cells per square millimeter, providing an objective, reproducible measure of immune infiltration.
Artificial Intelligence in Digital Pathology
Role of AI Digital pathology converts glass slides into high-resolution images. AI algorithms use machine learning—especially deep convolutional neural networks—to identify cell types, classify staining intensity, and count cells automatically.
Benefits for biomarker analysis:
- Speed: AI reduces time from minutes of manual review to seconds.
- Reproducibility: Automated thresholding and cell classification minimize observer variability.
- Sensitivity: AI consistently detects low-level biomarker expression, ensuring eligible patients are not missed.
Together, these advances support more accurate and efficient diagnostics, enhancing treatment planning for patients with lung cancer. AI-assisted PD-L1 scoring and TIL quantification promise to standardize critical biomarkers across pathology labs worldwide.