PD-L1 which is a co-regulatory molecule exhibits expression on the surface of different types of cells, such as immune cells as well as epithelial cells. By binding to its receptor PD-1 on lymphocytes, it generates an inhibitory signal toward the T-cell receptor (TCR)-mediated activation of lymphocytes [5]. Although PD-L1 might act as a tumor suppressor through inhibition of cancer stem cell (CSC) features in cholangiocarcinoma, tumor cell-intrinsic PD-L1 has an essential role in the promotion of cancer stemness, epithelial mesenchymal transition EMT, tumor invasion, and chemoresistance in many other tumor types. Essentially, stimulation of octamer-binding transcription factor 4 (OCT4) signaling and upregulation of EMT induce PD-L1 expression in malignant cells, thus recommending a potential immune evasion mechanism employed by CSCs throughout metastasis [6].
PD-L1 is expressed in tumor cells, and it has a critical function in tumor immune escape and in the development of a permissive immune microenvironment, via three or more mechanisms: (i) energizing tumor-reactive T cells via binding to its receptor PD-1; (ii) making tumor cells resistant to CD8+ T cell and Fas ligand-mediated lysis; and (iii) tolerizing T cells by reverse signaling via T cell-expressed CD80 [18, 19]. Moreover, PD-L1 expressed by tumor-accompanied immune cells is the main parameter responsible for tumor-accompanied immune deficiency [6].
The current study demonstrated that nearly 60% of cases showed positive PD-L1 expression in tumor cells (Fig. 1A–D). PD-L1 expression in HGSC was approached in only few studies. These studies found that PD-L1 expression was variable in cases of serous carcinoma ranging from 11 to 60% of the cases [16, 20,21,22].
In our study, there was no statistically significant correlation between PD-L1 expressions in tumor cells and the patient age, tumor stage, L.Ns, metastasis, or FIGO stage. These results are in agreement with a meta-analysis study conducted in 2019 and found that PD-L1 expression was not related to tumor grade, stage, L.N condition, or tumor histology [23]. In contrast to these results, Drakes et al. reported that low PD-L1-expressing cells in tumor tissue were significantly accompanied with progressive disease as well as high-grade tumors [17].
The potential causes for inconsistent outcomes may involve cancer type, tumor heterogeneity, sample size, clinical stage, several interventions, the time point of PD-L1 measurement, and the various methodologies utilized in the study (including recognition approaches and techniques) [6].
Also, we found no association between PD-L1 expressions in tumor cells and development of ascites, peritoneal deposits, or recurrence. But PD-L1 expression was higher among patients with residual tumor, 82.4% compared to 54.5% among patients with no residual tumor, with marginal statistical significance (p 0.07). In contrast, PD-L1 expression in tumor cells was accused to promote peritoneal dissemination by repressing cytotoxic T cells function in another study that found that patients with positive cytology had positive PD-L1 expression in tumor cells [5].
Expression of PD-L1 was not associated with receiving NAT (p= 1). But as known, chemotherapy induces local immune suppression in ovarian cancer via nuclear factor kappa-light-chain-enhancer of activated B cells (NF-kB)–mediated PD-L1 upregulation. Thus, an association between chemotherapy and immunotherapy targeting the PD-L1/PD-1 signaling axis might enhance the anti-tumor response and provide a talented novel therapeutic option against ovarian cancers [24]. Also, PD-L1 expression was higher among patients with high Ki-76 but without statistical significance. There was no any association with clinical markers CEA (p=1) or CA 125 (p=0.36).
Murakami et al. studied both intraepithelial and stromal CD8-positive lymphocytes and found them correlating significantly with immunoreactive type of ovarian HGSC [25]. Also, the results of Rojas and his work group found a statistically significant correlation between ovarian HGSC and CD8-positive intratumoral-infiltrating lymphocytes, but they did not establish a histopathological classification of those ovarian HGSC cases [26].
In our study, we found low CD8 expression in 34 cases of HGSC (63%), while high expression was found in 20 cases (37%) (Fig. 1F–H). Importantly, there was a negative correlation among high TILs CD8+ T cells and tumor PD-L1 expression with high statistical significance (p≤0.001). Such result is in accordance with a prior research which confirmed that expression of PD-L1 by tumor cells predicted paucity of intraepithelial TILs in ovarian malignant tumors [11].
There was no statistically significant correlation among CD8 expression and additional clinicopathological parameters. Such results are consistent with Adams et al.’s results who found no correlation of high intraepithelial CD8+ T cells with age, tumor stage, or tumor grade [10]. Also, we found that CD8 expression was not associated with Ki-76 expression, CEA, or CA 125. On the contrary, a previous result reported a significant positive correlation between expression of Ki67 and frequency of intraepithelial CD8+ T cells (p = .041). They hypothesized that mitotically active ovarian cancers, which are more likely to exhibit genetic instability, express a more diverse antigenic repertoire, such as neoantigens which elicit a cellular immune response [10].
High expression of PD-L1 was accompanied with considerably worse OS in cervical cancer, NSCLC, gastric carcinoma, esophageal carcinoma, glioma, and other malignant tumors. Conversely, the prognostic value of PD-L1 for particular types of malignant tumors remains debated. Particular researches recorded that high PD-L1 might predict promising outcomes [7]. Factors which affect the accurateness of PD-L1 immunohistochemistry staining are as follows. Types of antibodies utilized are different in various researches. The cutoff value of PD-L1 staining positivity is different. Also, PD-L1 expression in tumors is not uniform, and sampling time as well as site might affect the results of PD-L1 staining.
Importantly, DFS of patients was significantly lower among the group with positive expression of PD-L1 compared to the group with negative expression of PD-L1 (p 0.011). This result is consistent with a bioinformatics research which revealed that PD-L1 mRNA expression was closely accompanied with poor DFS [23]. However, that meta-analysis found that PD-L1 was not linked to DFS at the level of protein expression and suggested further studies on PD-L1 expression in tumor cells of ovarian HGSC. Thus, PD-L1 might become a talented therapeutic target for DFS of cases with this ovarian cancer. In contrast, only Webb et al. recorded that PD-L1 expression has a good prognostic effect on disease-specific survival in HGSC [16].
As regards PD-L1 and OS, we found that the mean survival time in patients with positive PD-L1 expression was less than patients with negative expression of the markers. However, this change was not statistically significant (p 0.089). Such result is in agreement with previous studies that found no correlation between the presence PD-L1 in tumor cells and OS in ovarian HGSC [17, 22, 27]. Also, a meta-analysis conducted after that suggested that PD-L1 expression was not related to OS [23].
However, few studies reported a significantly shorter OS in cases with PD-L1–positive epithelial ovarian tumors in comparison with PD-L1–negative tumors [5, 8, 10]. But in combining PD-L1 with CD8, the existence of PD-L1+ cells and CD8 TILs was associated with better prognosis than CD8 TILs alone [16]. The inconsistent outcomes of PD-L1 expression in ovarian HGSC might be owing to the various detection and scoring systems used to quantify PD-L1, differences in sample size, and temporal and spatial factors.
On the other hand, DFS of cases was not significantly different in the group with TILs CD8 expression >10% from the group with TILs CD8 expression <10% (p 0.739). But the mean OS time in cases with CD expression >10% was significantly higher than cases with CD expression <10% (p 0.043). These results are in agreement with a previous study which found that cases whose tumors had intraepithelial T cells experienced longer progression-free and OS as compared with patients whose tumors lacked intraepithelial T cells [11].
The association of antitumor immune response (intraepithelial T cells) with prolonged survival suggests that ovarian cancers are intrinsically immunogenic tumors [28]. Intraepithelial T cells and improved clinical outcome may be due to the direct function of tumor-infiltrating T cells or merely to a correlation of T cells with indolent tumors with low proliferation [10]. Optimizing approaches to select tumor-reactive TILs and expand them under optimum costimulation states which permit preferential expansion of specific T-cell phenotypes. Only a proportion of cases might be not eligible for TILs adoptive therapy, as tumors are either non-resectable or yield no tumor-reactive TILs [11].
Therapeutic potential of PD-L1 antibodies to reactivate antitumor immunity in ovarian malignant tumors is now highly considered [29]. Although clinical trials of PD-1 blockade in epithelial ovarian carcinoma are still at an initial stage, objective response rates have so far ranged from 6 to 17%. This recommends that further factors might require to be considered to more accurately predict responses [16].
The correlation of immune escape mechanisms with poor survival suggests that ovarian cancers can respond to identical immunotherapy procedures as patients with other immunogenic tumors [28]. So, antibodies which block PD-L1 or PD-1 can profoundly improve the efficacy of immune therapy.