Prognostic implications of PD-L1 expression in patients with soft tissue sarcoma

This study was conducted in a retrospective cohort of patients who were pathologically diagnosed with CD99 positive sarcomas such as rhabdomyosarcoma, synovial sarcoma, Ewing sarcoma, epithelioid sarcoma, and mesenchymal chondrosarcoma, between 1994 and 2013 at Yonsei Cancer Center. A total of 82 formalin-fixed, paraffin-embedded tissue blocks were available for examination of PD-L1 expression. All hematoxylin and eosin (H&E) slides were independently reviewed by two experienced pathologists (E.K.K. and S.H.K).

The clinicopathologic variables such as sex, age, maximal tumor size, tumor histology and grade, tumor location, tumor stage, initial presentation of disease, and the status of the resection margin were reviewed retrospectively based on electronic medical records. Tumors were graded according to the French Federation Nationale des Centre de Lutte Contre le Cancer (FNCLCC) criteria [8]. Staging was determined using the 7th edition of the American Joint Committee on Cancer guideline of tumor, node, and metastasis (TNM) classification. The study was approved by the institutional review board of Severance Hospital, Seoul, Korea.

Simple, inexpensive, and precise paraffin TMAs were constructed with a conventional microcompound table and a drill grinder. The original H&E slides were reviewed by two pathologists (E.K.K and S.H.K). One or two different tumor areas per case were selected for tissue microarray construction. Core tissue biopsies (3 mm in diameter) were taken from the individual paraffin blocks (donor blocks) and arranged in recipient paraffin blocks (tissue array blocks) using a trephine apparatus. All TMA blocks were confirmed by H&E staining.

Immunohistochemistry (IHC) staining of PD-L1 was performed on the TMA blocks using a mouse monoclonal antibody for PD-L1 (clone 130021, R&D Systems, Minneapolis, MN). Four-micrometer-thick sections were prepared and stained using a Ventana automatic immunostainer (Ventana, Benchmark, Tucson, AZ) [9]. After deparaffinization, heat-induced antigen retrieval was performed using pH 6.0 citrate buffer (CC1 protocol, Ventana, Tucson, AZ), and reactivity was detected using the Cell Mark detection kit (catalog no. 315-M-94). The sections were incubated with the primary antibody for 32 min at room temperature (dilution 1:100). The slides were counterstained with hematoxylin. Positivity for PD-L1 was evaluated and determined independently by two experienced pathologists (E.K.K. and S.H.K). In the case of discrepancy, a consensus was made through an in-depth discussion on multi-head microscopic observations. The intensity and percentage of PD-L1 positive tumor cells were counted manually in at least 100 viable tumor cells from 4 representative fields in each case. Because there was no consensus on the scoring system for PD-L1 expression, especially in sarcoma, we modified the previous protocol [10] and describe PD-L1 expression as much detail as possible by a semi-quantitative manner using the intensity multiplied by the proportion. The staining intensity was graded as negative (0), weak to moderate (1), or strong (2), and the proportion was categorized by the percentage of positive cells as 0, no positive tumor cells; 1+, less than 10 %; 2+, 10–50 %; and 3+, >50 %. Based on the multiplication score, a score of 0 or 1 was considered as negative PD-L1 expression, whereas scores greater than 2 were considered positive PD-L1 expression.

The Statistical Package for the Social Sciences (SPSS) version 18.0 (IBM, Chicago, IL) was used for statistical analyses. The correlations between PD-L1 expression and clinicopathologic variables were analyzed using the independent sample t-test for the continuous variables and the chi-square test for the discrete variables. For survival analysis, recurrence-free survival (RFS) was defined as the time interval between surgery and tumor recurrence or last follow-up. OS was defined as the time interval between the diagnosis of metastatic/recurrent disease and death or last follow-up. Survival analysis was performed using the Kaplan-Meier method with the log-rank test. Multivariate analyses for OS were performed with Cox’s regression. The accepted level of statistical significance was P < 0.05.

The baseline clinicopathologic characteristics of the patients are summarized in Table 1. The male-to-female ratio was 1.6:1, and the median age at the time of diagnosis was 26 years (range 1–78). The median tumor size was 5 cm, and two-thirds of cases showed low- or intermediate-grade tumors. Approximately half of the cases were rhabdomyosarcoma (n = 32, 39 %), followed by synovial sarcoma (n = 19, 23 %), Ewing sarcoma (n = 19, 22 %), epithelioid sarcoma (n = 7, 9 %), and mesenchymal chondrosarcoma (n = 6, 7 %). One third of the tumors were located in the extremities (n = 27, 33 %). Most patients had no distant metastasis at the time of diagnosis (82 %) and underwent surgical resection (93 %). Adjuvant chemotherapy was administered to 47 patients (57 %), and 37 of these patients (79 %) also received adjuvant radiotherapy.

Among the 82 sarcoma patients, PD-L1 was expressed in 35 cases (43 %). Representative images of PD-L1-positive and -negative staining for each histologic type are shown in Fig. 1. In the positive cases, PD-L1 was positive in the cytoplasmic membrane of the tumor cells and intratumoral endothelial cells. PD-L1 expression was significantly different according to the histologic subtype of sarcoma (P = 0.004). The proportion of PD-L1-expressing tumors was highest in epithelioid sarcoma (100 %, 7/7), followed by synovial sarcoma (53 %, 10/19), rhabdomyosarcoma (38 %, 12/32), and Ewing sarcoma (33 %, 6/18), whereas it was not expressed in mesenchymal chondrosarcoma (0 %, 0/6). Patients with PD-L1 expression had a more negative resection margin than the PD-L1-negative group (86 % vs. 62 %, P = 0.017). There was no significant difference in PD-L1 expression regarding age, sex, tumor size, location, histologic grade, surgical resection, and adjuvant treatment.

After a median follow-up duration of 33.8 months, 27 patients (33 %) had died at the time of survival analysis. The 5-year OS for all patients was 58.8 % and the 5-year OS rates of the subgroups were as follows: 64 % for rhabdomyosarcoma, 77 % for synovial sarcoma, 39 % for Ewing sarcoma, 18 % for epithelioid sarcoma, and 100 % for chondrosarcoma. There was a trend toward a worse RFS for the positive PD-L1 expression group compared to the negative group (Fig. 2a). Expression of PD-L1 in the tumor tissue significantly predicted shortened OS (5-year OS rate, 48 % vs. 68 %; hazard ratio [HR] = 2.545; 95 % confidence interval [CI] = 1.16–5.56; P = 0.015, Fig. 2b). Because the study cohort includes patients with initial distant metastasis and this could influence the survival outcome, we also analyzed the RFS and OS after excluding patients with metastatic disease. The trend for RFS and OS were almost comparable whether we exclude or not the metastatic patients (Additional file 1: Figure S1). The clinicopathologic variables significantly correlated with OS by univariate analysis were histologic subtype, initial metastasis, and PD-L1 expression (Table 2). In the Cox proportional hazard model adjusted for histologic subtype, initial metastasis, and PD-L1 expression, PD-L1 expression was significantly associated with shorter OS (P = 0.037, HR = 2.57, 95 % CI = 1.06–6.23).