Diagnostic biopsy does not accurately reflect the PD-L1 expression in triple-negative breast cancer

Abstract

AbstractPD-L1 expression is known to predict the benefits of immune checkpoint inhibitor therapy for triple-negative breast cancer (TNBC). We examined whether the PD-L1 expression evaluated in biopsy specimens accurately reflects its expression in the whole tumor. Immunohistochemistry was performed on 81 biopsy and resection specimens from patients with TNBC to determine their PD-L1 status. We found PD-L1-positive tumors in 23 (28%) biopsy specimens and primarily PD-L1-negative tumors in 58 (72%). The PD-L1 status was reevaluated in matching postoperative specimens of primarily PD-L1-negative tumors. Of them, 31% (18/58) were positive, whereas 69% (40/58) were negative. Considering the pre- and postoperative analyses, 41 (51%) patients had PD-L1-positive tumors, while 40 had PD-L1-negative tumors. We found 18 (22%) more PD-L1-positive tumors while examining the resection specimens compared to biopsies, and the difference was statistically significant (p = 0.0038). Diagnostic biopsies do not fully reflect the PD-L1 expression in TNBC. Our results suggest that a significant subset of TNBC patients may be misclassified as PD-L1-negative and disqualified from anti-PD-L1 therapy.