Targeting the Tumor Immune Microenvironment Could Become a Potential Therapeutic Modality for Invasive Pituitary Adenoma.

Objective: The study aimed to explore the relationship between the invasiveness and immune cell inltration in pituitary adenoma (PA) and provide the basis for immuno-targeting therapies. Methods: One hundred three patients who underwent surgery at a single institution were retrospectively identied. The inltration of macrophages and T lymphocytes was quantitatively assessed in PA. Results: The number of CD68 + macrophages was positively correlated with Knosp (P=0.003) and MMP-9 grades (P=0.00). The inltration of CD163 + macrophages differed among Knosp (P=0.022) and MMP-9 grades (P=0.04). CD8 + tumor-inltrating lymphocytes (TILs) were also positively associated with Knosp (P=0.002) and MMP-9 grades (P=0.01). Interestingly, MGMT expression was positively correlated with MMP-9 staining extent (P=0.000). The quantity of CD8 + TILs (P= 0.016), CD68 + macrophages (P=0.000) and CD163 + macrophages (P=0.043) were negatively associated with MGMT expression levels. The number of CD68 + macrophages in the PD-L1 negative group was signicantly more than in the PD-L1 positive group (P=0.01). The rate of PD-L1 positivity had positive correlations with the Ki-67 index (P=0.046) and p53 expression (P =0.029). Conclusion: Targeted therapy for macrophages and CD8 + TILs could be a helpful treatment in the future for invasive PA. Temozolomide (TMZ) may have better effects on the treatment of PA inltrating more immune cells. Anti-PD-L1 therapy may better respond to PA with higher Ki-67, p53 expression and more inltrating CD68 + macrophages. Multiple treatment modalities, especially combined immunotherapy, or immunotherapy with TMZ combination, could become a novel therapeutic strategy for invasive PA.


Introduction
Pituitary adenoma (PA) accounts for approximately 15% of intracranial tumors and represents the second most common primary brain tumor in humans [1]. Most PAs are noninvasive benign tumor growing slowly in the sellar and suprasellar regions [2]. However, about 35% of PAs are invasive adenomas usually in ltrating in the adjacent sphenoid sinus, cavernous sinus (CS) and internal carotid arteries (ICAs) [3]. In addition, third ventricular compression may lead to hydrocephalus. PA invades those important structures, make the goal of complete removal of tumor elusive, which further increases the risk of tumor recurrence [4]. As a result, local compression symptoms and neuroendocrine symptoms caused by invasive PA remain after traditional treatment. In addition, although not malignant, invasive adenomas often have aggressive biological behaviors [5]. Hence, treatment of invasive PA requires a multidisciplinary approach including surgery, radiotherapy, chemotherapy and even the very promising immunotherapy. It is delightful that signi cant progress has been made in immunotherapy.
Tumor-associated macrophages (TAMs) are an important constituent of the tumor microenvironment and are polarized to classically activated (M1) and alternatively activated macrophages (M2) characterized by CD163 positivity [6]. CD68 + is a pan-macrophage marker for both M1 and M2 macrophages. Some researchers have shown that M2 macrophages generally contribute to tumor initiation, progression, invasion, and angiogenesis and were correlated with poor prognosis of many cancers including glioblastoma (GBM) [7,8]. CD8 + tumor-in ltrating lymphocytes (TILs) have antitumor functions via initiating cytotoxic cascades [9]. Interestingly, CS invasion group of PAs tend to have higher CD8 + lymphocytes than CS non-invasion group [10]. In addition, PD-L1 expression is regarded as a crucial factor in predicting the effects of immune checkpoint inhibitors, which positively correlate with cavernous sinus invasion and increased TILs [10]. The anti-PD-L1 treatment has achieved an encouraging outcome in PA [11].
As a rst member of the matrix metalloproteinases (MMP) family to be linked to invasion of PA, MMP-9 could endorse angiogenesis and promote the invasion in many cancers [12,13]. Pleasingly, the MMP-9 inhibitor as a treatment of prolactinoma has made speci c progress on the rat model [14]. Temozolomide (TMZ) is an alkylating chemotherapeutic agent that has been used to against aggressive pituitary adenoma and carcinoma [15]. The presence of the MGMT (O6-methylguanine-DNA methyltransferase) results in inhibiting TMZ effectiveness, negative MGMT expression could predict responsiveness to therapy [16]. However, the association between the MGMT expression and invasiveness of PA remains controversial [17,18]. Consequently, it is necessary to understand comprehensively the effect of the immune microenvironment on tumor invasion to develop complete multidisciplinary treatments based on the assessment of imaging for invasive PA.
In this study, we mainly examined the expression of the extent of immune cells in ltration, immune checkpoint, MMP-9 and MGMT expression to investigate the relationship between these biomarkers and invasion and prognosis of PA. We try to explore the optimal and advanced multiple drug treatment strategy in invasive PA based on these results.

Patients and Design
This was a retrospective study of one hundred and three patients with PA who received surgery in SanBo Brain Hospital Capital Medical University between October 2017 and April 2018. The medical record, neuroimaging data, neuropathological material, and others were abstracted and collated from medical databases of the hospital. There were 54 males and 49 females, and the mean age of the patients was 45.02±14.91 years old (ranged from 15 to 79 years old) in the study.
There were seventy-nine patients with primary PAs and 24 patients with recurrent PAs. PA volume was estimated on the T1-enhanced MRI via the reported method [19]. The median tumor volume was 7.08 cm 3 (ranged from 0.28 to 245 cm 3 ). Ten patients with pituitary apoplexy were mainly diagnosed on MRI and clinical signs/ symptoms [20]. The Knosp classi cation grades was judged based on the neuroimaging diagnostic criteria [3]. Invasive and non-invasive PAs represented 47.57% (the 0 grade in 43 patients, 1 grade in 6 patients and 2 grades in 10 patients) and 52.43% (3 grades in 10 patients and 4 grades in 34 patients), respectively. According to 2017 WHO histopathological classi cation [21], PAs in this study were divided into 7 groups: 8 cases of somatotroph, 18 cases of lactotroph, one case of thyrotroph, 5 cases of corticotroph, 22 cases of gonadotroph, 46 cases of null cell and 4 cases of plurihormonal adenomas. The detailed patient information is shown in Table 1.

Immunohistochemistry
Immunohistochemical studies were performed on para n sections using an EnVision method. The PA specimens were xed with 10% neutral buffered formalin and further embedded in para n. Formalinxed and para n-embedded tumor specimens were cut into 4-µm thick sequential sections and processed for immunohistochemistry. The sections were then treated with 3% H2O2 for 5 min at room temperature to block endogenous peroxidase activity. The slides were blocked with 5% fetal bovine serum hours. After being washed with phosphate-buffered saline 3 times, the sections were stained with antimouse/rabbit polymer horse radish peroxidase-labeled secondary antibody (cat. no. PV-6000D; Zhongshan Bio-Tech Co., Ltd) for 30 min at 37˚C. Then, 3,3'diaminobenzidine (DAB) was applied for color development at room temperature for 5 min and sections were subsequently counterstained with hematoxylin. Each slide was individually reviewed and scored using Faramount mounting solution (Agilent Technologies Inc.). The Ki-67 value was assumed as the percentage of labeled nuclei in the PA specimen [21]. More than 10% of stained tumor cell nuclears were deemed as p53 positive expression [22].
The stained sections were viewed at low magni cation to identify 5-10 "hot-spots" areas with the highest density of stained immune cells. The number of selected "hot-spots" areas was mainly depending on the size of the specimen in microscopic examination. Every area was counted the number of stained immune cells under a 400-fold microscope. The average of the highest 3 values is used as the nal value. The method is a comprehensive reference to the previous studies [10,23].

Statistical analysis
The R language and GraphPad Prism software 9.0 (GraphPad Software, San Diego, CA, USA) were used to analyze the data and draw the mappings. Student t-test was conducted on two-group comparisons. Spearman correlations and regression analyses were mainly used to analyze the relationship between the factors. For the comparison of multiple groups, a one-way ANOVA was performed. The Chi-square test was used for categorical variables when needed. The Kaplan-Meier method was used to evaluate the overall survival (OS) and PFS. The univariate and multivariate analyses were performed by the Cox proportional hazards model. A P value < 0.05 was deemed signi cant.

MMP-9 Expression and Invasion
The  signi cantly increased risk of recurrence. K-M curve analysis showed that patients with incomplete resection of the tumor had low progression-free survival rates (P=0.01, Fig. 4). Patients with MMP-9 positive expression tended to have shorter progression-free survival than patients with MMP-9 negative (P=0.03, Fig. 4). However, strati ed by the extent of resection, and there was no statistically signi cant difference in progression-free survival between MMP-9 groups (P=0.10), PD-L1 groups (P=0.68) and MGMT groups (P=0. 44).

Discussion
The study demonstrated that the number of immune cells in ltrating PAs and MGMT expression, to some extent, were correlated with tumor invasiveness based on the images and molecular features. In addition, the positive rate of PD-L1 was positively associated with the Ki-67 index and p53 expression. Therefore, the above results reveal that there will be a novel and promising therapeutic approach for the invasive PA.
It is generally known that TAMs are crucial for the development and progression of many tumors [27]. Several studies have con rmed that macrophages, especially M2 macrophages, are associated with the tumor invasion and progression in PA [28,29]. The previous study demonstrated the number of CD68 + macrophages in ltrating in PAs was positively correlated with tumor size and Knosp grades for tumor invasiveness [23]. Some studies have supported MMP-9 overexpression was signi cantly associated with PAs' invasiveness [12,30], which was consistent with the result in this study. Our ndings con rmed that PAs' invasiveness based on neuroimaging and molecular pathology was correlated with in ltrating TAMs and M2 polarization. TAMs could enwrap tumor-associated blood vessels (TABVs) and create an environment conducive to tumor progression through promoting tumor angiogenesis and secreting growth factors [31]. It is generally known that increased angiogenesis serves a signi cant role in invasive or aggressive pituitary adenomas [32]. TAMs could secrete soluble proteases including MMP-9 to facilitate the in ltrative growth of TABVs through extracellular matrix (ECM) degradation and mobilize pro-angiogenic growth factors [33]. There are few studies on the topic, and further work is required to validate the speci c mechanism. Survival analysis showed that patients with positive MMP-9 expression had worse PFS and a higher risk of tumor recurrence. The higher level of MMP-9 was associated with a higher recurrence rate and a shorter recurrence-free interval in ACTH-secreting pituitary tumors in the previous study [34].In addition, invasion of the CS limits the possibility to resect a tumor completely [35].
Incomplete resection was a signi cant independent predictor of recurrence of tumor in multivariate analyses and patients with incomplete removal of the tumor had a signi cantly shorter PFS in this study.
Controversies exist regarding the associations between CD8 + TILs and PA's invasiveness or aggressiveness. It has been reported that T-lymphocyte (CD8, CD4, FOXP3) recruited by pituitary neuroendocrine tumors-derived chemokines determine aggressive behavior [36]. However, there is currently no evidence to suggest that CD8 + TILs are signi cantly associated with tumor size and Knosp classi cation grade [10,37]. Interestingly, it was noted that CD8 + lymphocyte cells were positively correlated with Knosp grades in the study. Our study revealed the number of CD8 + TILs was positively correlated with the expression of MMP-9. MMP-9 plays a critical role in CD8 + T cells in ltration into tissues and exerts a regulatory role on CD8 + T-cell activation [38,39]. However, the underlying association between the MMP-9 and CD8 + lymphocytes was not investigated in PA. Our results above revealed that PA with invasive behavior and/or MMP-9 expression could present a microenvironment highly in ltrated by CD8 + TILs which may exert a speci c effect on the invasion of PAs. Whether CD8 + TILs could express immunosuppression phenotype or not warrants further study.
The positive rate of PD-L1 in PAs was not the same among studies and remains controversial in tumor features [37,40]. There was a trend suggesting a higher positive rate of PD-L1 in non-functional or invasive PA [10,37]. We observed higher expression of PD-L1 occurred in null cell PA. In addition, it was reported that the PD-L1 expression is positively associated with increased CD8 + TILs [37]. However, the result was not seen in this study. Interestingly, CD68 + macrophages were more in ltrated in the positive PD-L1 group. This result suggests that CD68 + macrophages may promote the expression of PD-L1, further encouraging the formation of an immunosuppressive microenvironment. Pang et al. [41] demonstrated that TAMs might contribute to the immunosuppressive tumor environment by secreting cytokines or proteases inducing the overexpression of PD-L1 in cancer cells and activate related signal pathways to accelerate the tumor growth. In addition, it seems that patients with positive p53 expression and a high Ki-67 index would bene t most from an anti-PD-L1 therapy.
The lower the MGMT expression, the better response to the TMZ treatment in PA [42]. Our results suggested that there may be some interaction between the invasiveness and MGMT expression. A previous study showed that low-to-moderate MGMT immunoexpression was signi cantly more often in invasive PAs [43].However, although it was not signi cantly different between invasive and non-invasive PA in the imaging, the level of MGMT immunopositivity was positively associated with MMP-9 staining extent. The result may further exemplify invasive PA could be a suitable candidate for TMZ therapy, which is necessary to provide more evidence to document this. In addition, the expression of MGMT did show linear relationships with in ltrating immune cells, which made it possible to use immunotherapy in combination with TMZ to treat invasive PA.
Together, our results demonstrate that the tumor immune microenvironment consisting of immune cells and tumor cells serves a vital role in the invasion of PA. However, the effect of immunotherapy on pituitary adenomas has only been reported in cases, and the effect is uncertain. One ACTH-secreting PA patient progressed rapidly after four cycles of anti-PD-1 (pembrolizumab) [44]. It has been reported that clinical trials show that the therapeutic effect of a single immune checkpoint inhibitor is not satisfactory in glioblastoma (GBM), the attention has been shifted focus to immunotherapies combining other treatments, the novel approach might bene t more for patients with GBM [45]. In summary, we con rmed that the signi cant differences in the in ltration of immune cells (CD8+ TILs, CD68+ and M2 macrophages) and the expression of PD-L1 and MGMT between the invasive and noninvasive PA. Based on our results and combined with the experience of glioma immunotherapy, targeted macrophages / CD8 + TILs combined with PD-L1 or other types of immune checkpoints may provide a breakthrough in immunotherapy for invasive PA. Whether to add immunotherapy based on TMZ treatment may bene t patients more, which is expected to be further studies con rmed by preclinical and clinical trials.

Conclusions
Targeted therapy for immune cells could emerge as a potential treatment for invasive PA. High in ltrating immune cells in PA might indicate a better TMZ therapeutic effect. As for the PA with a higher Ki-67 index, p53 expression, and more CD68 + macrophages in ltration, PD-L1 inhibitors may be more effective. In addition, the incomplete removal of the tumor, recurrent PA and positive-MMP-9 expression are con rmed as increased risk of factors for recurrence in PA with surgery. Combined immunotherapy or immunotherapy with TMZ combination could open new chapters in the treatment of invasive PA.

Declarations
Funding The study was supported by the Beijing Municipal Science and Technology Commission of China (grant no. Z181100001718199).
Con icts of interest/Competing interests The authors declare that they have no con icts of interest.
Ethical approval This study was approved by the Ethics Committee of Sanbo Brain Hospital. Ethical authorization number is SBNK-YJYS-2019-002-01 Consent to participate Written informed consent for publication was obtained from all participants.

Consent for publication Not applicable
Availability of data and material The clinical datasets generated during and/or analyzed during the current study are available from the corresponding authors on reasonable request.
Authors' contributions SH, XLQ and CXY designed the study; ZCY, XMT and CW collected data and revised the manuscript for important intellectual content; ZCY, CW, KY, XLQ and LPZ performed the histological examination of the samples; ZCY, XMT and NL analyzed the data and ZCY wrote the article. All authors