Prostate cancer (PCa) is the most common cancer among men in developed countries. Although with the development of treatment methods, including chemotherapy, radiotherapy, hormones and surgical resection, the recurrence rate of PCa patients is still high [20]. Previous studies have shown that TME plays a vital role in the development, progression and recurrence of cancer [21]. Therefore, it is important to study the TME of PCa in this study to determine biomarkers that can predict DFS of PCa patients after RP.
In order to study the TME of PCa, we calculated the immune score, stromal score and estimate score of each PCa sample extracted from the TCGA database by applying an ESTIMATE algorithm. The results showed that higher immune, stromal, and estimate scores were associated with poorer DFS, higher Gleason score, and higher AJCC T stage in PCa patients with malignant tumors. Subsequently, we divided PCa patients into high/low immune score groups and high/low stromal score groups, and identified 515 cross-sectional DEGs. The GO and KEGG analyses of DEGs showed that DEGs mainly participated in TME, such as immune response, inflammatory response, cell adhesion, extracellular matrix organization, and leukocyte migration. These processes may inhibit tumor progression and metastasis, thereby improving DFS. We found that these DEGs have a strong correlation with the immune response and tumor immune microenvironment. In addition, we applied univariate COX and Lasso-Cox regression model to construct a RS model based on 18 DEGs which screened from 515 DEGs. In this model, DFS in the high-RS group was significantly lower than that in the low-risk group, so recurrence in PCa patients could be well predicted. The AUC of 1-year, 3-year and 5-year survival rates in the RS model were 0.778, 0.754 and 0.75, respectively. In addition, we found that the RS model constructed with 18 genes was more sensitive to prognosis than Gleason score. The stratified analysis showed that the RS model also had strong prognostic capability for PCa patients with negative surgical margins (R0).
Among them, the expression levels of C1QC, COL1A1, HOPX, ITGAX, STAB1, TGFB1 were low, and the prognosis was good. On the contrary, we found that the low expression levels of other genes were related to the poor prognosis of PCa patients.
C1QC belongs to C1Q and plays an important role in adaptive and innate immune responses. Studies have shown that C1QC can promote the adhesion, migration and proliferation of malignant pleural mesothelioma [22], and the increase in C1QC levels in patients with sarcoma is associated with poor prognosis [23]. The high expression of COL1A1 gene will cause unrestricted growth factors, which in turn will benefit tumor proliferation [24]. The HOPX gene may be involved in the malignant transformation of cancer cells. Studies have shown that higher HOPX expression is an independent adverse prognostic factor for acute myeloid leukemia [25]. ITGAX expression level is positively correlated with aggressive prostate cancer [26]. STAB1 is an identified oncogene whose increased expression promotes tumorigenesis and tumor progression [27], and it is associated with poor prognosis in many cancers. TGFB1 is often up-regulated in tumor cells and highly secreted into the prostate environment, partially mediating the immunosuppressive effect on NK cells and promoting the invasion and metastasis of prostate cancer [28].
APOF can act as a tumor suppressor for hepatocellular carcinoma, and the decreased expression of APOF is associated with poor prognosis [29]. The genetic variation of the nicotinic cholinergic receptor gene (CHRNS) may affect the risk of lung cancer [30]. The low expression of CLIC6 in breast cancer is related to high histological grade [31]. The tumor suppressor gene EGR-1 can directly mediate the apoptotic function through the transcriptional upregulation of Bax-mRNA and protein and the increase of oligomerization and activation [32]. FEV is rich in alanine c-terminal, indicating that it may act as a transcriptional inhibitor [33]. FOS is considered as a regulator of cell proliferation, differentiation, and transformation and participates in MAPK signaling pathway [34, 35]. GJB1 is abundantly expressed in other well-differentiated cell types such as prostate and pancreas. In prostate tumors, the ability to assemble GJS from GJB1 and GJA1 is impaired [36]. GNG2 is involved in the signal transduction of GPCR and CCR5 pathway in macrophages, and the expression level of GNG2 in malignant melanoma is decreases [37, 38]. The protein encoded by the HSD11b1 gene is a microsomal enzyme involved in the synthesis and regulation of prostaglandins. Up-regulation of OLFML3 enhances self-renewal of glioma stem cells and triggers primary tumor immunity PLTP plays a crucial role in mediating the association between triacyl lipid A and lipoprotein, which is beneficial to the anticancer properties[39]. As a candidate cancer suppressor, low TGM3 expression is associated with poor overall survival rate of neck cancer [40].
In conclusion, the RS prognostic model constructed by these 18 genes has not been reported and may be a new prognostic factor for PCa. Furthermore, multivariate COX model showed that RS and Gleason scores were two independent prognostic indicators. Based on RS and Gleason score, the nomogram for the prediction of DFS rate was established. The calibration diagram and AUC for the probability of recurrence at 1, 3 and 5 years showed that the prediction was consistent with the actual observation, and the prediction ability was strong. The main purpose of the nomogram is to quantify the risk of clinical events based on various predictive factors, provide personalized scores for each patient, and provide new ideas for personalized treatment of PCa patients.
Differences in the composition of Tiics subsets and immune checkpoints between high-RS and low-RS PCa patients were also explored. The concentration of M2 macrophages and Tregs was higher in the high-RS group. In contrast, the low-RS group had a higher proportion of CD8 T cells, resting memory CD4 T cells, and monocytes. The expressions levels of CTLA-4, PD-1, LAG-3, TIM-3 and TIGIT in high-RS patients were significantly higher than those in low-RS patients (P < 0.05). Previous studies have shown that resting memory CD4 T cells can further differentiate and have multiple functions, including restoring immune tolerance to autoantigens or heteroantigens and promoting CD8 T cells against tumors [41, 42]. Tregs expressing CTLA-4 play a crucial role in the maintenance of immunological self-tolerance and homeostasis and suppressing the anti-tumor immune response[43]. CTLA-4 is expressed in activated CD + 4 and CD + 8 T cells, which can terminate the response of activated T cells and mediate the inhibitory function of Tregs [44]. Overexpression of PD-1 on CD8 + T cells is an indicator of T cell depletion [45]. Inhibiting or knocking out LAG-3 will release the inhibitory function of Tregs on T cells. TIM-3 suppresses anti-tumor immunity by mediating T cell depletion. TIGIT can suppress immune cells in multiple steps of the tumor immune cycle [46]. In our study, the proportion of Tregs in high-RS patients was higher, the expression of immune checkpoint CTLA-4, PD-1, LAG-3, TIM-3 and TIGIT was higher, and the prognosis was poor, suggesting that the immunosuppressive environment and the high expression of immune checkpoint may be the reason for the poor prognosis of PCa. In addition, these results suggested that anti-CTLA4 drugs blocking immune checkpoint leads to T-cell activation, which is an ideal strategy for treating cancer. Anti-immune checkpoint antibody treatment will be more beneficial to high-risk PCa patients than low-risk patients, resulting in a better prognosis.
However, this study also has certain limitations. First, this study only conducted bioinformatics research on public databases, and the follow-up time of the validation set has not exceeded five years. Next, we should verify the results of this study through clinical patients in the prospective design. Secondly, 18 hub genes related to immune cell infiltration should be further studied to clarify the regulatory mechanism of PCa immune infiltration.