Accumulating evidence suggests that TME, including cancer cells and immune cells, supports tumor growth and evasion (19, 20). Several studies have pointed to the critical task carried out by CD39 in generating this immunosuppressive environment (21, 22). In this regard, we first confirmed that human lung cancer microenvironment was characterized by a population of CD8 + TILs expressing the ectonucleotidase CD39. Especially, the CD39 + CD8 + T cells was ubiquitously infiltrating in lung cancer tissue, irrelative of several clinic parameters, such as stage, age, pathologic types and lesion size, showed a tumor associated accumulation in this area.
In the present study, we further demonstrated that CD39 + CD8 + T cell subsets were selectively accumulated in peritumor rather than intratumor compartment, especially in larger tumors. Our analysis also showed that anti-tumor function of CD39 + CD8 + T cells was impaired by secreting less granzyme B. Recent developments in studies of tumor heterogeneity have provoked new thoughts on cancer management. For example, in melanoma patients treated with mitogen-activated protein kinase inhibitors, higher levels of CD8 + T cells from the intertumoral rather than peritumoral environment, along with lower levels of myeloid cells, have been reported to be associated with a favorable outcome (23). In estrogen receptor (ER)-negative/ human epidermal growth factor receptor 2 (Her2)-negative (24) and Her2-positive (25) breast cancer patients, a high degree of immune infiltration in tumor stroma was found to be associated with increased survival and complete response rates, respectively. Also, CD39highCD8 + T cells may contribute to the immunosuppressive tumor environment by promoting PD-L1 expression on tumors (26). Taken together with our results, we assumed that the characteristic spatial distribution may increase immunosuppressive activity of TILs derived from peritumoral environment, which facilitates the immune escape and promote tumor invasion.
As we know in the tumor microenvironment, prolonged or high-level expression of PD-1, Lag3, and CD244 (2B4) and other inhibitory receptors is a cardinal feature of exhausted T cells in both human disease and animal models (27, 28). Previous, virus-specific CD8 + T cells have been reported to express both CD39 and PD-1 in patients with chronic and resolved disease (29). Here, we found the percentage of infiltrated CD39 + CD8 + T cells showed a positively correlation with the PD-1 + CD8 + T cells, and CD39 + CD8 + T cells from both sites exhibited the higher percentage of PD-1 + cells compared to CD39- counterparts. These results further indicated that PD-1 expression on CD39 + CD8 + T cells identified it as an exhausted subset found in the lung cancer. Meanwhile, we confirmed that CD39 + CD8 + T cells derived from peritumor were more enriched with PD-1 in contrast to that from intratumor TME. Considering with our results that whose dominant distribution within peripheral compartment might represent an immune escape pathway, combining targeting CD39 and PD-1 may represent a novel immunotherapeutic approach.
Checkpoint blockade targeting PD-1 and PD-L1 interactions is by far the most effective immunotherapy approach, which, leading to durable complete responses for a subset of patients under a wide range of indications (30, 31). Although biomarker stratification is available, clinical responses vary (32, 33). Therefore, the search for novel biomarkers with improved response prediction is ongoing. Several studies have showed that CD8 proportion, in combination with tumor mutation burden and other signatures such as PD-L1, can predict the response of NSCLC patients to PD-1/PD-L1 blockade (34), especially the density of CD8 + T cells in the invasive tumor margin (23). Interesting, a previous study showed that patients with relatively higher proportion of CD39 + CD8 + T cells tend to have an improved response to PD-1/PD-L1 blockade therapy (35). And in our research, we found that peripheral PD-1 + CD39 + CD8 + T cell proportion had a stronger correlation with response status than CD39 + CD8 + T cell proportion, which might be a better biomarker to predict response to anti-PD-1 therapy. For this reason, we assumed that assessing the proportion of regional PD-1 + CD39 + CD8 + T cells in the tumor microenvironment can be a useful complementary approach to PD-L1 staining in patient stratification prior to immunotherapy.
Taken together, our findings have further deepened our knowledge of spatial heterogeneity of CD8 + T cell subsets, strongly emphasize PD-1 + CD39 + CD8 + T cell whose dominant distribution within peripheral compartment may facilitate invasion of cancer cells. In addition, we firstly described regional PD-1 + CD39 + CD8 + T cell as a candidate biomarker for immunotherapy to lung cancer, provide a new direction for developing the future immunotherapy approaches.