Combined PD-L1/CD8 immune phenotyping in breast cancer: exploratory associations across molecular subtypes and...

Programmed death-ligand 1 (PD-L1) expression and CD8+ tumor-infiltrating lymphocyte (TIL) density are established biomarkers of tumor-immune interaction in breast cancer; however, the biological and prognostic significance of their combined assessment across molecular subtypes remains incompletely defined. This study aimed to characterize combined PD-L1/CD8 immune phenotypes, evaluate their clinicopathological associations, and explore their relationship with survival outcomes in a molecularly heterogeneous breast cancer cohort.

This multicenter retrospective study included 137 patients with invasive breast carcinoma from two institutions. PD-L1 expression was evaluated separately in tumor cells (TCs) and tumor-infiltrating lymphocytes (TILs) by immunohistochemistry. CD8+ TIL density was also assessed immunohistochemically, and tumors were classified as CD8-high or CD8-low using a predefined threshold of ≥50%. Based on combined PD-L1 and CD8 status, cases were categorized into four immune phenotypes: Type I (PD-L1+/CD8-high), Type II (PD-L1+/CD8-low), Type III (PD-L1-/CD8-high), and Type IV (PD-L1-/CD8-low). Associations between immune phenotypes, clinicopathological characteristics, and survival outcomes were analyzed using appropriate statistical methods.

PD-L1 positivity was observed in 62 of 137 tumors (45.3%) and was significantly associated with higher histological grade (p = 0.020), but not with pathological tumor stage, clinical stage group, or lymph node status. A significant positive correlation was identified between PD-L1 expression in the immune-cell compartment and CD8+ TIL density (Spearman ρ = 0.453, p < 0.001), whereas no correlation was observed between tumor cell PD-L1 expression and CD8+ TIL density (ρ = -0.021, p = 0.812). In a hypothesis-generating subgroup comparison limited by small cell counts, CD8-high status was more frequent in Luminal B than in triple-negative breast cancer (36.0 vs. 13.7%; p = 0.036). The most common immune phenotype was Type IV (43.1%), whereas Type I accounted for 9.5% of cases. In exploratory analysis, patients with the Type I phenotype had the highest mortality rate (23.1%) and the shortest median follow-up-based survival estimate (37 months) among the four groups (log-rank p = 0.014); given the small number of events (11 deaths overall, 3 within the Type I subgroup) and the use of follow-up duration with cross-sectional mortality rather than fully annotated time-to-event data, these findings should be interpreted as exploratory associations rather than definitive prognostic evidence.

Combined PD-L1/CD8 immune phenotyping identifies biologically distinct immune microenvironmental states in breast cancer beyond conventional molecular subtype classification. While PD-L1 positivity was associated with high histological grade and immune-cell infiltration, the PD-L1+/CD8-high phenotype demonstrated the least favorable survival profile in this predominantly hormone receptor-positive cohort on unadjusted analysis, an association that was attenuated to borderline significance after adjustment for baseline metastatic burden. These findings support the concept that concurrent PD-L1 expression and abundant CD8+ T-cell infiltration may not invariably reflect effective antitumor immunity and may instead be consistent with an exhausted or functionally impaired immune state. If validated in larger cohorts with direct assessment of T-cell exhaustion markers, combined PD-L1/CD8 phenotyping may provide clinically relevant insights for immune stratification and patient selection in immunotherapy-oriented approaches.