*1.4. Immune Checkpoint Inhibitors and Their Potential Use in ACP*

Immune checkpoint inhibitors have shown promise in the treatment of a number of cancers. Specifically, inhibition of the programmed cell death protein (PD1) and its ligand (PD-L1) with the agents, nivolumab and pembrolizumab, has resulted in improved survival in cancers including melanoma and non-small cell lung adenocarcinoma [41,42]. The availability of these agents and their relatively favorable side effect profile has resulted in numerous studies investigating their efficacy in various cancers/tumor types.

PD-1 is an important protein involved in inhibitory immune signaling and is an essential regulator of the adaptive immune response [43]. In cancers, PD-1-expressing tumor-infiltrating T cells can be disabled by PD-L1 expressed on the surfaces of tumor cells themselves or alternatively by PD-L1 on the surface of other infiltrating immune cells. The binding of PD-1 to its ligand results in the suppression of the immune response to the cancer cells [43,44]. Checkpoint Inhibitors reverse this process and allow T cells to once again attack the cancer. Predicting the response of a particular tumor or cancer to PD1 inhibitors such as nivolumab and pembrolizumab is difficult and Taube et al. aimed to identify those factors that best predicted a robust and meaningful response to therapy [44]. In a prior study by the same group they found that anti-PD-1 therapy produced an objective response in 20–25% of patients with treatment-resistant NSCLC, renal cell carcinoma and melanoma and that PD-L1 expression by tumor cells seemed to be associated with a response to therapy [45]. In their follow up study they aimed to further investigate various factors that might predict a response to anti-PD-1 therapy including PD-L1 expression by tumor cells, PD-L1 expression by infiltrating immune cells, PD-L2 expression by tumor cells and other tumor microenvironment factors. They found that in their cohort only the expression of PD-L1 by tumor cells correlated with both an objective response (as defined by the "Response evaluation criteria in Solid Tumors" or RECIST criteria) and clinical benefit (*p* = 0.025 and 0.005 respectively). The correlation of the expression of PD-L1 by infiltrating immune cells with a clinical response did not reach statistical significance although the correlation with clinical benefit was statistically significant (*p* = 0.038). Of note, expression of the PD-1 receptor on tumor infiltrating lymphocytes (TILs), expression of PD-L2 by tumor cells or TILs, and other microenvironment immune factors did not correlate with outcomes. In addition, it is important to reiterate that even in those tumors expressing PD-L1 on tumor cells, only 39% of patients (9 out of 23) had an objective response [44]. That being said, these therapies have provided an option for patients with aggressive and treatment-resistant cancers for whom previously there were few if any good options. Due to the efficacy of immune checkpoint inhibitors more and more work

is being undertaken to identify other cancers and tumors that may be amenable to such therapy including craniopharyngiomas.

Recent work by Coy et al. [33] demonstrated the expression of PD-L1 in epithelial cells lining the cysts and intrinsic PD-1 expression in the beta-catenin over expressing whorl-like epithelial cell clusters in ACP. As previously discussed, these clusters are thought to play a pivotal role in tumor growth in ACP via a number of mechanisms [13,20,21], rendering targeting of PD-1 as an attractive potential therapy. Another study by Witt et al. [46] also demonstrated elevated PD-L1 expression in ACP. As mentioned above, numerous previous studies on other solid cancers have demonstrated that the expression of PD-L1 can be predictive of the response to the PD-1/PD-L1 inhibitors [44,45]. Again, as previously mentioned, such a finding far from guarantees a response and in these landmark papers they found that even in patients that expressed PD-L1 on tumor cells, the response rate to the treatment was only 39% [44]. In addition, Witt et al. [46] nicely demonstrated, using T cell exhaustion testing of various types of ependymomas, that elevated PD-L1 expression in tumors can be indicative of either tumor adaptations to hide from the innate immune response or due to normal T-cell antigen-activation, a known function of PD-1. In their study they utilized functional T cell exhaustion assays that stimulate T cells via exposure to Phorbol 12-myristate 13-acetate (PMA)/ionomycin. Subsequent to stimulations their group used a Milliplex Map Kit (Millipore) to measure the concentration of several cytokines including IFN-gamma. They found that infiltrating T-cells in RELA fusion supratentorial ependymoma did not secrete IFN-gamma. They concluded that this suggested that in the case of RELA fusion ependymoma, the increased expression of PD-1/PD-L1 results in the exhaustion of infiltrating T-cells and immune evasion by the tumor [46]. On the contrary, they found that in group B ependymomas (which also express high levels of PD-1), infiltrating T-cells were, in fact, capable of secreting IFN-gamma after stimulation with PMA/ionomycin. They posited that in these tumors, elevated expression of PD-1 was representative of normal T-cell activation in response to the tumor [46]. As such, although the findings by Coy et al. [33] of elevated PD-1/PD-L1 expression in ACP are exciting and may result in an alternative treatment strategy in resistant and multiply recurrent cases, further investigation is necessary to fully elucidate the implications of this increased PD-1/PD-L1 expression in ACP before any widespread implementation.
