Programmed Death-Ligand 1 (PD-L1) expression and clinical outcomes in oral squamous cell carcinoma: evidence from a...

Despite the recognized biological relevance of the PD-1/PD-L1 axis in immune evasion, the prognostic significance of PD-L1 expression in surgically treated oral squamous cell cancer (OSCC) patients remains largely unclear.

To investigate the prognostic significance of PD-L1 expression in a large cohort of OSCC patients primarily treated surgically in a single center.

This analysis included 156 consecutive OSCC patients treated with curative-intent surgery with their clinicopathologic variables, treatment details and clinical treatment outcomes. PD-L1 expression was categorized using the Combined Positive Score (CPS), defining PD-L1 positivity as CPS ≥ 10. Sensitivity analyses were repeated at the alternative thresholds CPS ≥ 1 and CPS ≥ 20. Survival and recurrence endpoints (OS, DFS, LRCR, NCR) were evaluated using Kaplan-Meier/log-rank methods and Cox regression with multivariable adjustment for key confounders.

Of 156 consecutive OSCC patients, 120 (76.9%) were PD-L1-positive. PD-L1 status was not associated with any clinicopathological parameter investigated herein. At 3 years, OS (76.1% vs. 68.7%), DFS (63.5% vs. 42.6%), LRCR (77.1% vs. 75.5%), and NCR (93.0% vs. 91.0%) were comparable between PD-L1-positive and PD-L1-negative patients, respectively. Neither binary PD-L1 classification nor continuous CPS was associated with OS, DFS, LRCR, or NCR in multivariable Cox regression, with no differential effect by disease stage. In adjuvant-treatment-stratified analysis, PD-L1 positivity was associated with improved DFS in patients without adjuvant therapy, but not in those who received it. In sensitivity analyses, findings at CPS ≥ 20 were concordant with the primary analysis, whereas at CPS ≥ 1 PD-L1 positivity was associated with superior OS (HR = 0.280 [95% CI: 0.106-0.745]; p = 0.011).

Within a mean follow-up of 15.7 ± 12.5 months, PD-L1 expression assessed by CPS was not independently associated with survival or recurrence outcomes in surgically treated OSCC at the prespecified cutoff of CPS ≥ 10, nor at CPS ≥ 20. An isolated overall survival association at CPS ≥ 1 was driven by a very small PD-L1-negative reference group and should not be regarded as robust. Taken together, these data argue against a reliable standalone prognostic value of PD-L1 in the early postoperative period while the limited observation time does not permit conclusions on late recurrence or long-term survival.