How to Determine PD-L1 Status
PD-L1 status is determined by performing immunohistochemistry (IHC) on formalin-fixed, paraffin-embedded (FFPE) tumor tissue using specific FDA-approved antibody clones, with the choice of assay dictated by which immune checkpoint inhibitor therapy is being considered.
Specimen Requirements
The testing requires FFPE tumor tissue specimens 1. While archived FFPE tissues are suitable for PD-L1 testing 2, fresh biopsies, core needle biopsies, fine-needle aspirates (FNAs), or surgical resection specimens can all be used 3. However, adequacy rates vary:
- Whole-block sections: 100% adequate
- Core biopsies: 92% adequate
- FNAs: 88% adequate 3
Critical caveat: Alcohol-based fixatives (like Cytolyt™) significantly reduce PD-L1 expression compared to formalin fixation, potentially misclassifying 25% of cases into different clinically relevant PD-L1 expression categories 3. Therefore, formalin fixation is strongly preferred.
Assay Selection Based on Therapy
The specific IHC assay must match the intended checkpoint inhibitor 1:
- Pembrolizumab: Requires PD-L1 IHC 22C3 assay kit
- Atezolizumab: Requires PD-L1 IHC 28-8 assay kit
- Other agents: SP142 and SP263 clones are also FDA-approved for specific therapies
These assays are not interchangeable despite similar performance characteristics 4, 5. Each uses different antibody clones, protocol conditions, instrumentation, and scoring systems 4.
Scoring Systems
Scoring methods differ critically between assays 1:
28-8 Antibody (Atezolizumab)
Uses a 3-tiered system based on tumor cell staining:
- 1% cutoff
- 5% cutoff
- 10% cutoff
22C3 Antibody (Pembrolizumab)
Uses a different 3-tiered system:
- <1% (negative)
- 1-49% (low positive)
- ≥50% (high positive)
The scoring is based on Tumor Proportion Score (TPS): the percentage of viable tumor cells showing membranous PD-L1 staining of any intensity 6, 7.
Technical Considerations and Pitfalls
Tumor Heterogeneity
PD-L1 expression shows marked spatial heterogeneity 8:
- In 15.2% of needle biopsy-sized specimens, TPS varied by >10% at different tissue depths
- In 42.4% of cases, TPS at different depths fell into different clinical categories (<1%, 1-49%, ≥50%), potentially altering treatment decisions 8
Practical implication: When PD-L1 expression falls near clinical cutoffs (especially 1% and 50%), there is higher risk of discordance between sampling sites 3.
Quality Control Requirements
The assays are highly sensitive to protocol deviations 4. Slight variations increase the risk of misclassifying patient PD-L1 tumor status 4. Rigorous validation is mandatory, particularly for non-formalin-fixed specimens 7.
Interobserver Variability
Significant variability exists, particularly in the 1-49% PD-L1 expression range 7. This intermediate range poses the greatest challenge for reproducible interpretation.
Cytohistological Concordance
When using cytology specimens, concordance with histology varies 7:
- At 1% cutoff: 54-100% concordance
- At 50% cutoff: 82-100% concordance
Standardization Requirements
Per Society for Immunotherapy of Cancer guidelines, PD-L1 expression testing has undergone significant standardization and validation 2. However, the lack of a universal biomarker and scoring system across different antibodies remains a major limitation 1.
Each laboratory must:
- Validate their specific assay platform 2, 4
- Use appropriate positive and negative controls
- Follow manufacturer protocols precisely 4
- Participate in external quality assessment programs when available 7
Clinical Context
PD-L1 expression correlates with tumor characteristics 6:
- Higher in squamous cell carcinoma vs. adenocarcinoma
- Higher in poorly differentiated vs. well-differentiated tumors
- Associated with male gender
While PD-L1 expression is informative for predicting checkpoint inhibitor response, it cannot unequivocally predict clinical outcome 2. Expression data are most valuable when combined with tumor mutational burden and immune cell infiltrate data 2.