With the success of immunotherapy in various solid cancers, the expression of programmed death-ligand 1 (PD-L1) in cancer cells has been attracting attention. The aim of this study was to analyze PD-L1 expression in resected UPS and evaluate its clinical impact.
All consecutive patients who were treated with UPS in Asan Medical Center, between January 1995 and December 2016, were retrospectively reviewed. PD-L1 staining was performed using SP263 antibody on their archived tissues. PD-L1 H-score was calculated by intensity (0-3) multiplied by proportion (0-100).
A total of 176 patients with localized UPS who underwent curative-intent surgical resection were included in this study. There were 104 males (59.1%) and 72 females (40.9%), and the median age of patients was 59 years at the time of diagnosis. The most frequently involved primary site was extremities (84/176, 47.7%), followed by abdomen and pelvis (46/176, 26.1%), thorax (29/176, 16.5%), and head and neck (17/176, 9.7%). The PD-L1 analysis was available on slides from 114 (64.8%) patients and PD-L1 expression was observed in 83 (72.8%) cases. Intensity of staining varied between weak (44/83, 53.0%), intermediate (29/83, 34.9%), strong (10/83, 12.0%). Distribution of PD-L1 expression using different cut-off values were as follows: ≥1%: 83/114 (72.8%); ≥5%: 79/114 (69.3%); ≥10%: 52/114 (45.6%); ≥50%: 20/114 (17.5%). The 5-year disease free survival (DFS) rate seemed to be better in patients with PD-L1 positive than in patients with PD-L1 negative (56.5% vs 41.9%), but no statistical significant was observed (p = 0.183). There was a trend for better 5-year DFS rate with higher proportion of PD-L1 or higher PD-L1 H-score (proportion: 0% vs 1-49% vs ≥ 50, 41.9% vs 52.9% vs 70.1%, p = 0.098; H-score: 0 vs 1-100 vs ≥ 101, 41.9% vs 53.1% vs 75.7%, p = 0.087) without statistical singnificance.
In this study, PD-L1 protein expression was seen in substantial cases of resected UPS and it appeared to be a positive prognostic biomarker for longer DFS, although not statistically demonstrated. Further studies are needed comprehensively understand the impact of immune repertoire in UPS.
Clinical trial identification
Legal entity responsible for the study
Has not received any funding.
All authors have declared no conflicts of interest.