Thyroid cancer is the most common malignancy of the endocrine system. In addition to ATC, the majority of patients with thyroid cancer have a better prognosis and higher overall survival. The treatment for patients with thyroid cancer is usually based on surgical resection and also includes radioiodine therapy and long-term thyroid hormone replacement therapy[25, 26]. It is well known that autophagy is a double-edged sword, which can both promote and inhibit tumorigenesis. Studies have shown autophagy is related to the progression of thyroid cancer. The induction of autophagy can make thyroid tumor cells sensitive to certain therapies and plays an important function in the regulation of apoptosis. With the evolution of high-throughput sequencing technology, an increasing number of studies have shown that lncRNAs play an important role in the occurrence and development of cancer. Nowadays, survival models for autophagy-related genes have been established in thyroid cancer, but no model for autophagy-related lncRNA survival is available. Therefore, it is necessary to construct a risk signature based on autophagy-related lncRNA in thyroid cancer.
In this study, the TCGA database was used to collect RNA sequences and related clinical information of thyroid cancer patients, and the HADb database was used to collect autophagy-related gene information. After univariate Cox regression analysis and multivariate Cox regression analysis, a nine-autophagy-related lncRNAs significantly associated with prognosis was identified. We stratify patients with thyroid cancer into high-risk and low-risk groups based on the median prognostic risk scores, and the results showed that the high-risk group had a poorer prognosis. The ROC curve and AUC validated the accuracy of the risk model. The AUC of the risk score was 0.905, proving that the accuracy of risk signature was superior.
Of the nine autophagy-related lncRNAs, five lncRNAs (AL136366.1, AC008063.1, AC092279.1, DOCK9-DT, LINC02471) were protective factors for the prognosis of thyroid cancer, while other four lncRNAs (LINC02454, AC096677.1, AC004918.3, AL162231.2) behaved the opposite. LINC02454 and AL136366.1 have been shown to be pivotal mediators in the pathophysiological process of PTC and are closely associated with the development and prognosis of thyroid cancer. Tan et al. reported that LINC02454 is highly expressed in PTC. And LINC02454 may function as an oncogene that inhibits the apoptosis and enhances proliferation of PTC cells. Zhang et al. constructed an overall survival model composed of eight signature lncRNAs. It showed that DOCK9-DT is a protective factor for thyroid patients. Dong et al. pointed out AC004918.3 was involved in constituting a prognostic signature for clear cell renal cell carcinoma (ccRCC) and could be a valid prognostic indicator for ccRCC. Chen et al. demonstrated that LINC02471 can promote the development of PTC. Knockdown of LINC02471 can also inhibit the invasion and metastasis of PTC, and promote apoptosis of PTC cells by directly targeting miR-375. Cal et al. reported that LINC02471 can be used as a molecular biomarker for the progression and prognosis of thyroid cancer. Yang et al. indicated that LINC02471 may be a valid prognostic indicator for ccRCC. These results are almost same as ours and further proving that our results are reliable.
To further explore the functions of nine autophagy-related lncRNAs signature in thyroid cancer, we performed KEGG pathway analysis and GO enrichment analysis. KEGG pathway analysis indicated that the pathways association with spliceosome, RNA polymerase and base excision repair were prominently enriched in the low-risk group. In addition, in GO enrichment analysis we found that differentially expressed lncRNAs are involved in the synthesis of U2-type spliceosomal complex, RNA splicing, SNRNA binding, viral transcription and the initiation of RNA polymerase. These results contribute to our further understanding of the nine autophagy-related lncRNAs signature.
To sum up, we constructed and confirmed a nine autophagy-related lncRNAs risk model which can perform as an independent prognostic factor of thyroid cancer. These may provide a certain theoretical basis for the screening, diagnosis and treatment of patients with thyroid cancer in the future.
This study is the first to construct a prognostic risk model of autophagy-related lncRNAs using bioinformatics in thyroid cancer. Even though we used the extremely authoritative TCGA database, there were still problems with our study. We only used the TCGA database, which lacks further validation of external datasets. Furthermore, we need some systematic and comprehensive in vivo or in vitro experiments to further validate our conclusions.