In this prospective exploratory study, we comprehensively assessed multiple PBMC subsets in patients with NSCLC who were receiving ICI therapy. We found that the patients who developed irAEs had increased CD86+pDC levels at baseline compared with patients who did not develop irAEs. Furthermore, CD86+pDC levels in patients with irAEs were significantly decreased by weeks 2 and 8 after initiation of anti-PD-1 therapy. The CD86+mDC levels were not associated with irAEs. Additionally, increased proportion of NK cells in PBMCs were associated with the efficacy of ICI. Our data indicate that the assessment of PBMC subsets may aid in predicting the efficacy and safety of ICI therapy.
DCs are a group of antigen-presenting cells that stimulate T cells via co-stimulatory factors, including CD86 (also known as B7-2) [7][8]. Among several subsets of DCs, pDCs produce type I interferons (IFN-I) in response to viruses, unlike mDCs which mainly act as conventional antigen-presenting cells [9]. Although pDCs constitute only 0.2–0.8% of PBMCs, emerging evidence suggests that pDCs contribute to the pathogenesis of autoimmune diseases including systemic lupus erythematosus (SLE), systemic sclerosis, psoriasis, autoimmune thyroid diseases, type I diabetes, autoimmune pancreatitis, and inflammatory bowel diseases (IBDs), as well as having a role in immune responses to infectious diseases [10][11][12][13][14][15][16]. For example, antibody-mediated depletion of pDCs reduced type I interferon responses and disease activity in patients with cutaneous lupus [10]. In patients with IBDs, increased numbers of pDCs were observed in inflamed mucosa, and increased levels of peripheral CD86+pDCs were associated with disease activity [14]. Furthermore, patients with type I diabetes had significantly higher pDC levels in their peripheral blood than healthy individuals [17]. Interestingly, autoimmune thyroid disease, IBDs, and type I diabetes are known to occur after ICI therapy. Given the autoimmune aspect of irAEs, it is reasonable to propose that an increase in activated pDCs at baseline may affect the development of irAEs.
The decrease in CD86+pDC levels after ICI therapy in patients with irAEs showed another interesting similarity with autoimmune diseases. In patients with type I diabetes, the increased number of peripheral blood pDCs at diagnosis tended to decrease after 2 years, whereas the number of peripheral pDCs was stable over the 2 years in control subjects [17]. Additionally, in patients with SLE, psoriasis and autoimmune thyroid diseases, pDC levels were decreased in peripheral blood rather than increased [13][15][18]. This seemingly paradoxical behavior is thought to be from an accumulation of pDCs in tissue lesions. In fact, increased pDCs have been observed in involved organs (such as skin, lymph nodes, kidney, or thyroid) in patients with SLE, psoriasis, or autoimmune thyroid diseases [13][15][19]. The decrease in circulating CD86+pDCs after ICI therapy possibly reflects the recruitment of these cells from the peripheral blood to specific organs.
The accumulation of pDCs in lesional tissues is a common characteristic in several autoimmune diseases; however, controversy exists regarding the levels of circulating pDCs, which were increased in type I diabetes and IBDs, and decreased in SLE, psoriasis, or autoimmune thyroid diseases [13] [14][15][17][19]. Although the precise mechanisms involved are unknown, the timing of the evaluation during the course of disease (at diagnosis or after progression) or the specific organs involved may be associated with the differences in circulating pDCs among autoimmune diseases.
The pathogenesis of irAEs and differences in disease pathogenesis among different irAEs are largely unknown. Shared antigens between tumors and organs, inflammation generated by cytokines and activated immune cells, or pre-existing organ inflammation (i.e. autoimmune diseases) might be potential mechanisms [5]. The irAEs occur in a wide variety of organs, and therefore, organ-specific mechanisms may exist for each irAE. However, some common mechanisms might also be involved because all the irAEs were caused by autoimmunity triggered by ICIs. As well as having a role in innate immunity by producing IFN-I, pDCs promote the differentiation and maintenance of autoreactive B cells [20][21]. Given their multiple roles in innate and adaptive immunity, pDCs might be involved in various irAEs. However, there were no obvious differences in the baseline levels of CD86+pDCs among the different types of irAEs. Furthermore, decreased proportion of CD86+pDCs were observed after ICI therapy regardless of the type of irAE. These results indicated that pDCs were associated with a common mechanism across irAEs. However, the current study evaluated a limited number and type of irAEs. The organ-specific roles of pDCs should be investigated further.
The association between increased peripheral blood NK cell levels and ICI efficacy has been previously reported and was also observed in the current study [3][24]. In animal models, depletion of NK cells attenuated the anti-tumor effects induced by PD-1/PD-L1 blockade. Additionally, PD-1/PD-L1 signaling resulted in reduced anti-tumor activity of NK cells, which was restored after PD-1/PD-L1 blockade [25]. The underlying mechanisms are largely unknown; however, these results indicate the considerable importance of NK cells in ICI therapy as a predictive factor of efficacy or therapeutic target.
The current study has four main limitations. First, this was an exploratory study with a limited number of subjects. We evaluated multiple surface markers on immune cells, which may have introduced possible false positives. Furthermore, PBMCs contained a small fraction of pDCs, and therefore, the results of this study should be interpreted with caution. To validate the utility of CD86+pDC levels for the prediction of irAE, further studies with larger cohorts will be required. Second, the association between CD86+pDC levels and severity or time to onset of irAEs was not evaluated. The differences in pDC function among the organs involved in the irAEs are also unknown. ICI-induced irAEs are a heterogeneous group of conditions that occur in various organs, and the roles of pDCs may differ according to the irAE phenotype. Third, the differences in distribution of immune cells between peripheral blood and organs were not determined. Although circulating CD86+pDC levels were not associated with the efficacy of ICI treatment, tumor-infiltrating pDCs likely play an essential role in cancer immunity. In animal models, tumor-infiltrating pDCs induced antitumor immunity by activating NK cells, conventional DCs, and CD8+T cells [26]. In the tumor microenvironment, tumor-infiltrating pDCs’ anti-tumor activity is attenuated by tumor-derived soluble factors, including transforming growth factor-β, which results in immune tolerance to the tumor [27]. Fourth, we only evaluated anti-PD-1 therapy. Several single or combinatorial ICI therapeutic strategies exist, including anti-PD-ligand(L)1 or anti-cytotoxic T-lymphocyte antigen-4 (CTLA-4) antibody therapy, combinations of ICI with chemotherapy, and combinations of ICI antibody therapies. Immune status may differ according to the type of immunotherapy used. In fact, the combination of anti-PD-1 and anti-CTLA-4 therapeutics demonstrated frequent occurrence of irAEs [28][29]. CTLA-4 suppresses T cell activity by binding to CD86; therefore, the CD86+pDC behavior during anti-CTLA-4 therapy is of interest. Immunotherapies will become more diversified in the near future as new immunomodulating agents emerge. The associations between host immune status and immunotherapy efficacy and safety should be investigated further.