Tumor-infiltrating lymphocytes (TILs) constitute a polyclonal lymphocyte population that migrates from the bloodstream to the tumor and target tumor associated antigens, playing, therefore, a key role in the anti-tumor immune response. Numerous studies have demonstrated the association of the density and distribution of TILs with patient response to chemotherapy and their overall prognosis in several solid tumors. Denkert et al. demonstrated that TILs are a favorable prognostic factor in HER2-positive breast cancer and triple negative breast cancer [14], while recently, TILs have emerged as a potential biomarker in highly aggressive tumors, such as melanoma [15].

ACC has been traditionally described as an immune deplete tumor, a fact mainly attributed to adrenal glucocorticoid production. In a pan-cancer analysis utilizing mRNA expression of immune-related genes (including PD-L1) in a TCGA (The Cancer Genome Atlas Program) cohort, Pare et al. showed that immune infiltration in ACC is indeed lower compared to other types of cancer [16]. In agreement with this data, an immune genomic analysis of multiple cancer types revealed that ACC, along with uveal melanoma, displayed the lowest leucocyte fraction (calculated based on data of DNA methylation) [17]. TCGA’s integrative genomic analysis of ACC revealed three distinct molecular subtypes (CoC, or cluster of cluster I, II, and III). The expression of immune-related genes is, in general, low, except for the CoC I subgroup, where immune infiltration is observed [18].

However, expression of immune-related genes, such as ERN1 and CEP55, points out the important role of TME in ACC [19]. An increasing number of studies in recent years highlight the significance of TILs as well as other types of immune cells in ACC patients’ prognosis. A recent study using immunofluorescence to quantify infiltrating cell populations in ACC samples, demonstrated that >80% of the samples were infiltrated by lymphocytes with CD8+ cells being the predominant type [20]. It should be noted, however, that the median number of infiltrating immune cells was rather low, which explains the lower quantitative immune gene-related mRNA expression observed in previous studies [16]. Tian et al. applied gene analysis and CIBERSORT algorithms—a computational method for quantifying cell fractions—to a TGCA cohort and showed that the immune infiltrate mainly consisted of T-cells, natural killer cells, mast cells, and macrophages while infiltration levels by each subtype showed strong correlation with the other subtypes. The number of mast cells correlated with survival, as well as with specific changes in signaling pathways [21]. Both the abovementioned studies proved a significant correlation of TIL levels with overall survival (OS) and recurrence-free survival (RFS). Increased levels of TILs were associated with lower pTNM and AJCC staging [21], and notably, metastatic foci seem to be relatively immune deplete compared to primary tumors. This suggests that immune evasion plays a crucial role in disease progression. The association of TILs, and more specifically CD8+ lymphocytes, with prognosis has been demonstrated in pediatric ACC patients as well [22]. Overall, current data suggest that, although lymphocyte infiltration is present in ACC and correlates with patient prognosis, it is quantitatively smaller comparing to other types of cancers.

In addition to the levels of infiltration, the pattern of distribution of infiltrating lymphocytes within the tumor could potentially play a significant role in the dynamics of the TME. The diffuse pattern of CD8+ lymphocytes seems to predominate in ACC compared to the multifocal pattern, which is dominant in benign adenomas [22].

Further investigation regarding the role of immune cells, other than the lymphocytes, in the microenvironment of ACC is warranted. Mast cells constitute a potential target population as they have been positively correlated with prognosis and their presence might exert a regulatory effect on lymph cell populations. Furthermore, emerging data highlight the significance of neutrophils in the microenvironment of many solid tumors, rendering them a potential target in immunotherapy. In a multicenter study, Bagante et al. reported that increased preoperative neutrophil/lymphocyte ratio (NLR) and platelet/lymphocyte ratio (PLR) are associated with decreased post-operative RFS in patients with ACC [23]. Neutrophils are responsible for releasing several chemokines such as TNF-a and VEGF, which are implicated in angiogenesis, steroidogenesis, tumor growth, and metastasis, while, at the same time, interfering with normal T-cell response. It should be noted that the elevated NLR was significantly associated with tumor secretion of glucocorticoids, suggesting that corticosteroid-induced lymphocyte depletion could also be contributing to the observed increased ratio.

Over the previous years, increasing importance has been attributed to the role of adipose stem cells in the TME and their association with increased tumor growth and invasiveness. The correlation of certain malignancies with obesity is well established. The crosstalk between adipose stem cells (ASC) and cancer cells seems to be mediated through the release of several growth factors (VEGF, PDGF, TGF-β), cytokines (IL-6, Il-8, IFN-γ, TNF-a, CXCL2, CXC12), and leptin by the former. Furthermore, several in vitro studies have highlighted the increased expression of matrix metalloproteinases in ASC and cancer cell cocultures, which could be pivotal in a tumor’s potential for migration and invasion ^[37][38][37,38].

Recently, Armignacco et al. performed similar co-culture experiments using human ASC cells and ACC cells (H295R cell line). According to this study, cancer cells showed increased proliferation and invasiveness when cocultured with ASCs in comparison to cancer cell cultures alone. On the other hand, ASCs showed decreased maturation and lipid content in the presence of H295R cells. The chemokine and molecular profile of the cocultures was investigated further, and it was demonstrated that the cocultures showed increased levels of CXC12 and CXC7-chemokines consistently associated with increased tumor migration. In addition, increased levels of leptin and IL-8 seemed to be implicated in the more invasive phenotype of the cancer cells in the coculture, compared to the control [39]. Indeed, several studies in the past have associated leptin with tumor aggressiveness in other solid tumors. Strong et al. reported increased proliferation and metastatic dynamics in ER-positive breast cancer in response to leptin produced from ASCs isolated from obese women [40], while Chen et al. reached similar conclusions regarding proliferation in ovarian cancer cell lines [41].

Approximately 60% of ACC patients present with hormonal excess syndromes. Among those, the majority present with either Cushing’s syndrome alone, or a mixed virilization-Cushing’s syndrome, inducing endogenous hypercortisolism. Furthermore, exogenous glucocorticoids are administered to many patients following adrenalectomy or during mitotane treatment. Even in the absence of high serum concentration, intratumoral glucocorticoid concentrations may be high due to activation of steroid synthesis pathway in some tumors [42]. Indeed, CoC II and III molecular subtypes have a steroid-high phenotype and render a worse prognosis, whereas CoC I is characterized by a significant up-regulation of genes in immune-mediated pathways [43].

The various mechanisms by which glucocorticoids impair immunity have been described in detail in numerous studies over the years. In regard to the T-cell response, which has a cardinal role in orchestrating anti-tumor immunity, glucocorticoids deplete T-cells by directly inducing apoptosis, downregulating IL-2 production, and inhibiting their release from lymphoid organs [44]. They also hinder T-cell antitumor function. It is obvious that successful ICI therapy requires intact cytotoxic T-lymphocytes. In addition, high glucocorticoids levels promote tumor cells’ survival and proliferation in vivo [45].

It is of no surprise, therefore, that adrenocortical tumors with glucocorticoid hypersecretion display diminished numbers of TILs, with CD3+ CD4+ being the predominantly affected subset. Landwehr et al. analyzed 146 ACCs and demonstrated a negative correlation between hypercortisolism and overall survival, with immune depleted cortisol-secreting tumors displaying the lowest survival rates (compared to non-immune depleted cortisol-secreting, non-immune depleted non-secreting, and immune depleted non-secreting tumors) [20]. Other studies have demonstrated that glucocorticoids interfere with anti-tumor immune response by downregulating MHCII and TLR-4 ^[46][47][46,47].

In support of the unfavorable glucocorticoid milieu, pembrolizumab administration to a patient with metastatic ACC, Lynch syndrome, and tumor-associated Cushing’s syndrome led rapidly to disease progression (PD) [48]. On the contrary, an impressive response to the drug was reported in another patient with normal cortisol levels, due to concomitant mitotane treatment [49]. Therefore, suppression of glucocorticoids’ negative effect on the immune system seems important to produce meaningful clinical benefit of IO in ACC patients.

Mitotane is effective in controlling glucocorticoid excess by inhibiting glucocorticoid biosynthesis and inducing increased steroid clearance and cortisol-binding globulin, but it takes several weeks [7]. In the case of severe Cushing’s syndrome requiring a rapid control, other agents can be used that inhibit adrenal steroidogenesis. Metyrapone is a steroidogenesis enzyme blocker that can be used concomitantly to chemotherapy and mitotane during the first weeks of treatment until mitotane therapeutic levels are obtained, leading to the rapid resolution of symptoms [50]. Another steroidogenesis inhibitor, which is also administered in patients with Cushing’s syndrome, is ketoconazole. It is less effective than metyrapone though, and it cannot be used in combination with mitotane due to hepatotoxicity. It is also effective in androgen excess. Finally, mifepristone is a synthetic steroid with progesterone receptor antagonist activity and glucocorticoid receptor antagonist activity at higher doses [51] that could also be used. Improvement in 66% of patients during the first month was observed in an international retrospective study [52]. However, there are concerns about mineralocorticoid adverse effects requiring monitoring. Apart from mitotane, all these drugs could be used in combination with IO to overcome hypercortisolism in ACC.

The immune modulatory role of androgens is extremely complicated and remains to be fully elucidated. There are reports in the literature that DHEA-S counterbalances the effects of glucocorticoids by interfering with the function of 11β-HSD1, which is responsible for cortisone to cortisol conversion [53]. It has also been suggested that it can potentially augment T-cell response by upregulating Il-2 production, and therefore boost T-cell proliferation, as well as potentiate T-cell mediated cytotoxicity [54]. Canning et al. have also documented a possible positive net effect of DHEA-S on the maturation of dendritic cells [55]. On the other hand, DHEA-S has been reported to downregulate the production of several inflammatory cytokines, such as TNF-a and IL-6 [56].

Further research, therefore, is warranted to clarify the role of DHEA-S in the glucocorticoid-rich microenvironment of adrenal cancer. It should also be noted that pure androgen hypersecreting ACC is extremely rare, as, in most cases, androgen hypersecretion is accompanied by Cushing’s syndrome. Identifying, therefore, the role of androgen hypersecretion on survival could be confounded by the increased cortisol concentrations in the same patients.

The landscape of genomic alterations of ACC is complex. Activation of WNT/β-catenin signaling has been recognized as an oncogenic driver in a large subset of ACC patients, mainly in CoC II and CoC III tumors [16]. β-catenin regulates development and homeostasis in different tissues. Preclinical data in melanoma shed light on the effects of WNT/β-catenin pathway activation on tumor immune microenvironment (TIME), characterized by a reduced production of some chemokines, such as CCL4 and subsequent reduced BATF3 dendritic cells. As a result, decreased T-cells recruitment and a defective effector T-cell trafficking into the TME were observed, preventing anti-tumor immunity [57]. Similarly, limited clinical data in human ovarian carcinoma and adenoid cystic carcinomas revealed an association of WNT signaling activation and lack of T-cell infiltration [58]. According to a pan-cancer integrative genomic analysis using the TCGA, activation of WNT/β-catenin signaling was enriched in non-T-cell-inflamed tumors, whereas ACC demonstrated the strongest correlation [59]. In silico analysis of the TCGA ACC tumors revealed that high Catenin β 1B (CTNNB1) gene expression correlated with cortisol excess, worsened survival, and decreased immunity, as manifested by fewer TILs [60]. The different elements of the TIME are depicted in Figure 1.

Given the immune exclusion caused by WNT/β-catenin activation, drugs inhibiting the pathway seem rational combination partners for future immune checkpoint inhibitor (ICI) trials. Targeting WNT/ β-catenin signaling is challenging, since it has a critical role in normal tissues homeostasis like bones and various physiological processes. There are no approved drugs in oncology. However, several effectors and inhibitors of WNT/β-catenin pathway have been tested at the preclinical level, as well as in early phase clinical trials of solid tumors and hematologic malignancies [61]. As an example, Porcupine inhibitors act by blocking the secretion of WNT ligands. The frequent deletions or loss of function mutations in the ZNRF3 gene, a gene which fosters turnover of cell surface Frizzled receptors of WNT, may render ACC harboring these alterations sensitive to these drugs. OMP18RS (vandictumab) is a monoclonal antibody against the Frizzled receptor that has been tested in a phase 1 study of breast pancreatic cancer with promising activity [62], whereas a combination phase 1 study with chemotherapy in pancreatic cancer was terminated due to bone-related adverse events [63]. PRI-724 is a small molecule inhibitor of WNT/β-catenin/CBP (a transcriptional coactivator) signaling, and was evaluated in early phase trials of pancreatic cancer [64].

There are no clinical data on these agents in ACC patients. In vitro studies in ACC models demonstrated that β-catenin targeting inhibited cell proliferation ^[65][66][65,66]. It should be noted that the WNT pathway cross talks with the Notch and Sonic Hedgehog pathways, and thus it is possible that multiple targeting is necessary.

Another important proportion of ACC displays a spectrum of mutations in the p53/Rb pathway, which have been correlated to aggressive histotype and bad prognosis ^[67][68][67,68]. These mutations have also been described as pathogenic in other cancers and represent the most relevant prognostic biomarker in ACC. TP53 is a tumor suppressor gene encoding for p53 protein, an important regulator of cellular stress response. The majority of TP53 mutations lead to stabilization and accumulation of the mutant protein, which is protected from ubiquitin-mediated degradation [69]. TP53 mutations that were identified in ACC result in either a negative function or absence of protein expression. Germline TP53 mutations are associated with pediatric ACC, as a component of LFS, whereas somatic mutations are observed mainly in adult cases [70].

TP53 inactivating mutations have been shown to confer an immunosuppressive phenotype, and, due to decreased MHC-I presentation, profound changes in chemokine/cytokine secretion increasing the recruitment and activity of myeloid and Treg cells [71]. Therefore, effector T-cells recruitment in the tumor is suppressed, promoting immune evasion. In the same direction, TP53 mutations were recently found to downregulate immune-related genes in hepatocellular carcinoma [72]. TP53 mutations and inactivation have also been reported in cancer-associated fibroblasts (CAF), affecting immune cell composition and function in TIME and leading to pro-inflammatory molecule production [73].

Strategies aiming at restoring p53 function in tumors and CAF could inverse immunosuppressive TME, which is observed in TP53-mutated tumors. Pharmacological p53 activation, viral vector-mediated p53 reintroduction, and restoration gene therapy have not provided satisfactory results so far [73]. Given the broad spectrum of TP53 mutations, mutant-specific reactivating drugs are probably required. Another approach to target p53 is the disruption of the chaperone machinery involved in its impaired degradation.

Additional genetic alterations have been described in different subsets of ACC patients, such as the dysregulation of the mTOR pathway [74]. Substantial evidence supports that the PI3K/AΚΤ/mTOR cascade plays a central role in immune cells’ homeostasis and activation. It can regulate chemokine-mediated immune cells, with diverse roles from promoting T-cell accumulation to immune evasion [75]. The PI3K/AΚΤ/mTOR pathway has been shown to regulate PD-L1 in different cancers, Treg cells, and myeloid-derived suppressor cells infiltration [76]. Therefore, further investigation is required to determine the role of mTOR inhibitors like everolimus and temsirolimus in the immune modulation of selected ACC subgroups. The mTOR pathway has already been suggested as a potential therapeutic target for ACC [77]. Although in vitro testing of these drugs led to the inhibition of ACC cells’ proliferation [78], preliminary clinical data failed to demonstrate efficacy as monotherapy [79].

Combination of IO with drugs targeting the above pathways is a promising strategy to overcome immune exclusion in ACC and warrants extensive research. Furthermore, injection of mature dendritic cells into tumors with β-catenin alterations could also be explored.