In advanced or metastatic triple‑negative breast cancer, how should PD‑L1 expression be assessed and which immunotherapy regimens are indicated based on PD‑L1 positivity?

PD-L1 Testing and Immunotherapy in Advanced Triple-Negative Breast Cancer

All patients with locally recurrent unresectable or metastatic triple-negative breast cancer who are candidates for immunotherapy must undergo PD-L1 testing with an FDA-approved assay before treatment initiation. 1

Mandatory PD-L1 Testing Requirements

Testing is required—not optional—for any patient being considered for immune checkpoint inhibitor therapy in the metastatic setting. 1 The 2022 ASCO guideline provides a strong recommendation (evidence-based, benefits outweigh harms) that PD-L1 expression testing on tumor and immune cells must be performed to determine eligibility for pembrolizumab plus chemotherapy. 1

Which Assay to Use

The choice of PD-L1 assay depends on which immunotherapy regimen you plan to use:

• For pembrolizumab: Use the 22C3 pharmDx assay on the Dako Autostainer Link 48 platform, which measures the Combined Positive Score (CPS). 1, 2, 3
• For atezolizumab: Use the VENTANA PD-L1 SP142 assay on the Ventana BenchMark Ultra platform, which measures the Immune Cell (IC) score. 4, 2, 5

Critical caveat: These assays are not interchangeable—each must be used on its designated platform, and CPS versus IC scoring identifies different patient cohorts. 3 A 2024 reproducibility study demonstrated almost perfect inter-rater agreement for CPS scoring (κ = 0.938 for 22C3, κ = 0.972 for SP263), but emphasized that CPS and IC identify distinct populations and cannot be substituted for one another. 3

Scoring Thresholds and Clinical Cutoffs

For pembrolizumab eligibility:

• CPS ≥ 10 is the FDA-approved threshold for first-line pembrolizumab plus chemotherapy. 1, 2
• At this cutoff, the KEYNOTE-355 trial demonstrated median PFS of 9.7 months versus 5.6 months with chemotherapy alone (HR 0.65,95% CI 0.49–0.86, p = 0.0012). 1
• Median OS improved to 25 months versus 15.5 months in PD-L1–positive patients. 4

For atezolizumab eligibility:

• IC ≥ 1% (≥1% of tumor area occupied by PD-L1–positive immune cells) is the threshold for atezolizumab plus nab-paclitaxel. 4, 2, 5
• The IMpassion130 trial showed median OS of 25 versus 18 months in PD-L1–positive patients receiving atezolizumab plus nab-paclitaxel. 4

Recommended Immunotherapy Regimens Based on PD-L1 Status

First-Line Treatment for PD-L1–Positive Disease

If CPS ≥ 10: Administer pembrolizumab 200 mg IV every 3 weeks plus chemotherapy (nab-paclitaxel, paclitaxel, or gemcitabine-carboplatin). 1, 2 This regimen achieved statistically significant PFS and OS benefits in the KEYNOTE-355 trial. 1

If IC ≥ 1% by SP142: Administer atezolizumab plus nab-paclitaxel. 4, 2 The 2021 ASCO guideline gives this a strong recommendation (evidence-based, benefits outweigh harms, moderate evidence quality). 4

Important nuance: The KEYNOTE-355 trial also evaluated patients with CPS ≥ 1, showing median PFS of 7.6 versus 5.6 months (HR 0.74), but this did not meet the prespecified statistical threshold (α = 0.00111) for significance. 1 Therefore, CPS ≥ 10 remains the validated cutoff for pembrolizumab approval, not CPS ≥ 1. 1

PD-L1–Negative Disease

When PD-L1 is negative or immunotherapy is not indicated: Use single-agent taxane chemotherapy (weekly paclitaxel is preferred) rather than combination chemotherapy. 2 Meta-analyses confirm that PD-L1–negative patients derive no significant OS or PFS benefit from immune checkpoint inhibitors. 6, 7

Alternative for PD-L1–negative metastatic disease: Nab-paclitaxel plus carboplatin yields superior outcomes (median PFS 8.3 months, OS 16.8 months) and is an acceptable platinum-based option. 2

Critical Testing Considerations and Pitfalls

Primary Tumor vs. Metastatic Tissue

Always test the metastatic lesion when feasible, not just the primary tumor. 8 A 2023 study found poor concordance (κ = -0.03) in PD-L1 status between primary and metastatic tumors using SP142, and moderate concordance (κ = 0.50) with 22C3. 8 Among patients who were PD-L1–negative in primary tumors by both assays, 23.3% converted to PD-L1–positive in metastases. 8 This means some patients initially deemed ineligible for immunotherapy based on primary tumor testing may actually qualify if their metastatic tissue is tested. 8

Inter-Assay Concordance

The concordance between 22C3 (CPS ≥ 10) and SP142 (IC ≥ 1%) at diagnostic cutoffs is substantial in primary tumors (κ = 0.80) but only moderate in metastatic tumors (κ = 0.60). 8 Do not assume that a negative result on one assay predicts negativity on the other—if clinical suspicion is high, consider testing with the alternate assay/platform combination. 3, 8

Quality Assurance Requirements

Pathologists performing PD-L1 scoring should be certified PD-L1 trainers or have specific expertise in breast pathology. 3 In-house assay validation, prediagnostic training, and participation in internal and external quality assurance programs are essential. 5

Predictive Value of PD-L1 Expression

Meta-analyses confirm that PD-L1 expression reliably predicts:

• Objective response rate: OR 1.78 (95% CI 1.45–2.19, p < 0.00001) for PD-L1–positive versus negative patients. 6
• 2-year OS: OR 2.47 (95% CI 1.30–4.69, p = 0.006) favoring PD-L1–positive patients. 6, 7
• 1-year PFS: OR 1.39 (95% CI 1.04–1.85, p = 0.02), with better predictive ability when immunotherapy is administered first-line (OR 1.38, p = 0.04) and combined with chemotherapy (OR 1.38, p = 0.03). 6

The predictive value is strongest for pembrolizumab in the first-line setting combined with chemotherapy. 6

Additional Molecular Testing

Beyond PD-L1, perform:

• Germline BRCA1/2 testing in all triple-negative patients to identify candidates for PARP inhibitors (olaparib 300 mg twice daily). 2
• Ki-67 proliferation index for risk stratification. 2

Toxicity Profile

Immune-related adverse events of any grade occur in approximately 26% of patients receiving pembrolizumab-chemotherapy versus 6% with chemotherapy alone. 1 Grade ≥3 immune-related adverse events occur in 5% versus 0%, respectively. 1 Thyroid disease affects roughly 23% of patients on atezolizumab. 4 Treatment discontinuation due to adverse events occurs in 16% with atezolizumab versus 8% with chemotherapy alone. 4