Effect of Tissue Sample Type on The Evaluation of PD-L1 (SP142) Expression in Breast Cancer

Background/Aim: PD-L1 expression is a key biomarker for immune checkpoint inhibitor therapy in breast cancer treatment. However, many factors affect PD-L1 assessment. This study evaluated how specimen-related factors affect PD-L1 expression in breast cancer and provides a guide for optimal specimen selection.

Patients and Methods: We retrospectively analyzed 30 consecutive breast cancer specimens submitted for PD-L1 testing at the Osaka University Hospital between November 2019 and November 2021. PD-L1 expression was evaluated using the VENTANA SP142 immunohistochemistry assay. Expression in tumor-infiltrating immune cells (IC) was quantified as the percentage of PD-L1-positive IC within the tumor area, and positivity was defined as IC ≥1%.

Results: The overall PD-L1 positivity rate was 47%. PD-L1 expression was significantly higher in surgical specimens compared with biopsy samples (65% vs. 23%, p=0.032) and in primary tumors compared with metastatic/recurrent sites (58% vs. 0%, p=0.018). All specimens from patients with Stage IV or recurrent disease were PD-L1-negative, compared with 74% positivity at earlier stages (p<0.001). Although not statistically significant, specimens from patients who had received chemotherapy within 40 days showed lower PD-L1 positivity than chemotherapy-naïve specimens (14% vs. 57%, p=0.086). In advanced disease (Stage IV/recurrent), 82% of specimens were from biopsies and 55% from metastatic sites, potentially explaining the lower PD-L1 positivity.

Conclusion: For optimal PD-L1 assessment in breast cancer, surgical specimens from primary tumors without prior therapy are preferable due to larger evaluable tumor areas. For patients requiring neoadjuvant chemotherapy or with de novo Stage IV disease, multiple biopsies of primary tumors using thick needles before treatment, with attention being paid to sampling tumor margins to account for potential immune-excluded phenotypes, are recommended.