It has been suggested that lncRNAs play a key role in the adjustment of the immune system and the tumor microenvironment[10]. Immune-related lncRNAs are prognostic markers of various types of cancer[16] linked to immune cell infiltration and are a potential target for cancer treatment[17]. lncRNA SNHG14 has been shown to be involved in the activation of the JAK-STAT pathway in cervical tumor cells [18]. In cervical cancer HeLa cells, the STAT3 binding sequence in the enhanced region of lncRNA MALAT1 proved to be the key to the activation of the MALAT1 promoter induced by IL-6 or STAT3 [19]. In the eleven immune-related lncRNAs of the risk model, STXBP5-AS1 targets miR-96-5p/PTEN axis to drive cervical cancer cell proliferation and invasion[20], and DLEU1 promotes cervical cancer cell proliferation and invasion via the miR-381/HOXA13 axis[21].
We used the AIC value to get the best cutoff point for the model fitting. The prognostic model performed better in distinguishing between risk groups of high and low. Patients with low-risk showed a favorable prognosis, which suggested that our model may be able to stratify risk. GSEA revealed that immune-related pathways were significantly enriched in the low-risk group. We also analyzed the tumor immune infiltration and efficacy of chemotherapy drugs in the treatment of CSCC, which determined that this model worked well. In this study, consensus clustering based on NMF was used to define CSCC subtypes. The CSCC sample could be separated into two subtypes with distinct molecular profiles with differences in immune cell infiltration and checkpoint-related biomarkers. A nomogram was established to show the exact agreement between the observation rate and the predicted rate of 1-year, 3-year, and 5-year OS. The nomogram had better prediction accuracy and higher clinical efficacy. passed calibration curve and DCA verification.
The ability of immunotherapy depends on the immunogenicity of the tumor microenvironment, so the understanding of the tumor microenvironment is the key to assessing the possibility of immunotherapy[22]. Immune cell infiltration can be a predictive biomarker of cancer immunotherapy [23]. Patients with more CD8 + T cell infiltration get a better response from pembrolizumab than those with less infiltration [24]. Macrophages suppressed T cell recruitment and regulated other aspects of tumor immunity, they can be regulators of tumor immunity and immunotherapy[25]. We used seven accepted methods to estimate the correlation between risk scores and immune cells, the methods include XCELL[26], QUANTISEQ[27], MCPcounter[28], EPIC[29], TIMER[30, 31], CIBERSORT[32], and CIBERSORT-ABS[33]. By integrating analysis, our results revealed that the risk model was negatively associated with macrophages M1, macrophages M2, and CD8 + T cells, meanwhile CD8 + T cells were significantly higher in the low-risk group. We can assume that patients in the low-risk group may be more suitable for immunotherapy.
Cancer cells activate inhibitory immune checkpoint pathways to prevent the occurrence of autoimmunity[34], therefore, suppression of the immune checkpoint approach is a promising treatment. Evidence suggests that immunotherapy is a novel therapeutic strategy for the treatment of cervical cancer[35]. Anti–CTLA-4, anti-PD-1, and anti-PD-L1 therapies are possible treatment options for cervical cancer, studies show[36–38]. Furthermore, in a well-defined clinical study of 115 cervical cancer patients, PD-L1 expressed in 19% of cervical tumors, and more than 50% of tumor-infiltrating CD8 + T cells expressed PD-1[39]. We explored the relationship between several common immune checkpoint proteins in the treatment of cervical cancer and risk models, showing that all six immune checkpoints were higher in the group of low-risk. Thus under this model, it suggested that patients with low-risk may benefit more from immunotherapy. The predictions of the TIDE algorithm indicated that patients with low-risk subtypes had a better response to immunotherapy, which was consistent with the above-mentioned studies. Our risk model also showed that low-risk patients were more sensitive to cisplatin, a small-molecule platinum compound that treat recurrent cervical cancer effectively[40], so we can formulate chemotherapy regimens for low-risk patients.
In our risk model, the risk scores are related to the clinical stage, and the risk scores of early cervical cancer are lower than that of the advanced stage. From the joint analysis above, We can assume that early-stage patients may be better suited for immunotherapy and appropriate chemotherapy.