We are pleased to know that our work has inspired Thelen and colleagues (1) to study CD39+CD8+ T cells in other human tumor entities. In accordance with our observations in patients with breast cancer (2), they found an accumulation of CD39+CD8+ T cells in head and neck squamous cell carcinoma, renal cell carcinoma, non–small cell lung cancer, and gastric adenocarcinoma. In this context, it is important to highlight another recently published article in which Simoni and colleagues found that CD8+ T cells specific for tumor neoantigens show high CD39 expression (3). As Thelen and colleagues' analysis was made in total CD3+ T cells, it would be interesting to study the frequency of CD39+ cells gating on CD8+ T cells to establish whether there is a correlation between the presence of CD39+CD8+ T cells and tumor immunogenicity. This analysis is based on the fact that all these tumor entities display different frequencies of somatic mutations and therefore different immunogenicities (4). Using RNA-seq studies, Simoni and colleagues also showed that CD39+CD8+ T cells exhibit an exhausted signature in lung and colorectal cancers, which supports our conclusions in breast cancer. Altogether, these studies are relevant in the design of checkpoints blockade therapies, which are being applied mainly in immunogenic tumors (5).
On the other hand, we agree that it is important to evaluate healthy tissues to exclude the possibility that the expression of CD39 on CD8+ T cells may be due to other tissue-related factors regardless of tumor presence. In this sense, we have analyzed samples from six primary breast cancer tumors matched with adjacent nontumoral mammary tissues. In the nontumoral tissues, a 6.9% ± 1.03% of CD8+ T cells expressed CD39, which was significantly lower than the 16.8% ± 5.28% of tumor-infiltrating CD39+CD8+ T cells, thus confirming the role of tumoral tissue in promoting the accumulation of this cell subset. In nontumoral mammary tissues and in nonmetastatic draining lymph nodes, the small population of CD39+CD8+ T cells also exhibited lower capacity to produce IFNγ and TNF than CD39−CD8+ T cells. This new data opens a new angle of study, in which it could be relevant to determine whether tumor-infiltrating CD39+CD8+ T cells are expanded from the healthy tissues or recruited from blood and then acquire their specific phenotype in the tumor microenvironment, as it has previously been described for Tregs in breast cancer (6).
Our study and the data provided by Thelen and colleagues have reinforced the hypothesis that CD39 may emerge as a new immune checkpoint in CD8+ T-cell–mediated antitumor response.
See the original Letter to the Editor, p. 5173
Disclosure of Potential Conflicts of Interest
No potential conflicts of interest were disclosed.
The research projects are supported by the following grants: MINCyT (PICT 2012-0480 and PICT 2015-1954 to C.L. Montes), SECYT 2012-2016 (C.L. Montes), Labex DCBIOL (ANR-10-IDEX-0001-02 PSL and ANR-11-LABX0043 to E. Piaggio and N. Núñez).