Identification of a novel prognostic gene signature in high-grade lung neuroendocrine tumors

doi:10.21203/rs.3.rs-2345264/v1

Download PDF

Research Article

Identification of a novel prognostic gene signature in high-grade lung neuroendocrine tumors

https://doi.org/10.21203/rs.3.rs-2345264/v1

This work is licensed under a CC BY 4.0 License

You are reading this latest preprint version

Background

The prognosis of high-grade LNETs patients was poor. Although previous studies have tried to explore molecular markers to stratify LNETs and to seek the therapeutic targets, there was little improvement in the treatment strategy and overall survival of high-grade LNET patients.

Materials and Methods

In our study, gene expression data and clinical features of LNETs patients were extracted from GSE30219 dataset, which was further divided into training cohort and validation cohort. Univariate cox regression analysis and LASSO method were used to construct a prognostic risk model. The predictive value of prognostic gene signature was evaluated by ROC curve analysis.

Results

An eight-gene signature predicting survival was constructed in training cohort and was further validated in validation cohort. This gene signature performed well in both early-stage and advanced-stage high-grade LNET patients. Moreover, multivariate cox regression analysis revealed that this gene signature was the only independent risk factor in high-grade LNET patients, while age, gender, pathological stage, T stage, N stage and M stage were not. GO analysis and KEGG pathway analysis of DEGs and GSEA analysis based on subgroups stratified by gene signature in high-grade LNET patients revealed the high activity of tumor cell division, proliferation and metabolization.

Conclusion

We identified a novel and reliable gene signature risk model for predicting prognosis in high-grade LNET patients, which could stratify high-risk LNETs patients and might provide new targets.

high-grade lung neuroendocrine tumor

overall survival

gene signature

risk model

Lung cancer continues as the predominant incident cancer and the leading cause of cancer-related death worldwide with about 2.1 million new cases and 1.8 million deaths each year(1, 2). Lung neuroendocrine tumors (LNETs) is a subgroup of lung cancers which possess neuroendocrine (NE) morphology and express NE markers such as neural cell adhesion protein 1, neuron-specific enolase, chromogranin A, and synaptophysin(3). According to the World Health Organization (WHO) 2015 classification, LNETs were divided into four histologic variants: typical carcinoid (TC), atypical carcinoid (AC), large cell neuroendocrine carcinoma (LCNEC) and small cell lung carcinoma (SCLC)(4). Typical and atypical carcinoid are well-differentiated and have a more favorable prognosis compared to LCNEC and SCLC(5). However, no prognostic difference was noted between LCNEC and SCLC in several clinical trials(6, 7). The analysis of telomerase activity suggests that LCNEC is more similar to SCLC rather than NSCLC(8). Correspondingly, TC patients are treated with surgery, AC and LCNEC patients are treated with multimodal approaches involving surgery and/or systemic therapy, while SCLC patients usually undergo systemic therapy(9). The underlying molecular distinction between those totally different biological behaviors of the tumors remains to be clarified. The next generation sequencing showed that the most malignant LNETs have developed from pre-existing carcinoids through sequential acquisition of different gene alterations(10). Previous studies have tried to explore molecular markers, including genes and pathway identification, to stratify LNETs and to seek the therapeutic targets(10–15). However, there have been little improvement in the fundamental treatment of LCNEC and SCLC(7, 16). Therefore, there is an urgent need to develop novel, practical therapeutic approaches for LNETs, especially LCNEC and SCLC. And a better understanding of the genomic changes occurring in LNETs is essential to identify new therapeutic targets and markers.

In our study, by assessing the global gene expression profile of LNETs, we constructed and validated a gene signature, which could help to clarify the potential mechanism between genes or molecules and the prognosis of LNETs patients, to identify high-risk LNETs patients and to provide new targets.

Data source and preprocessing

The RNA sequencing data and clinical information of LNETs patients were extracted from GSE30219 dataset, which was downloaded from Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo). The GSE30219 RNA sequencing data was first log2 transformed and quantile normalized, then genes detected with more than one probe were calculated by mean expression. 104 LNETs patients, including 24 carcinoids, 59 LCNEC and 21 SCLC, were enrolled in our study.

Because the data were obtained from GSE30219 dataset, approval for our study by the ethics committee was not necessary.

Identification Of Differentially Expressed Genes (Degs) And Enrichment Analysis

Due to the prominent prognostic discrimination between low-grade LNETs (carcinoids) and high-grade LNETs (LCNEC and SCLC), we sought to explore the DEGs between two types of LNETs. 80 high-grade LNETs patients were randomly divided into training cohort and validation cohort, each 40 patients. DEGs generated from the analysis of mRNA expression distinction between 24 low-grade LNETs and 40 high-grade LNETs of training cohort by using ‘edgeR’ package in R with P value < 0.05 and FDR < 0.05.

Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were performed by R package of ‘org.Hs.eg.db’ to clarify the function and pathway of the significant DEGs with P value of < 0.05 being statistically significant.

Construction And Validation Of Prognostic Degs Risk Model

314 DEGs from training cohort were screened by univariate cox regression analysis. Next, we used the least absolute shrinkage and selection operator (LASSO) method(17) for variable selection in a cox regression model to determine significant prognostic genes. The risk score formula was calculated by taking into account the expression of each gene and corresponding regression coefficient as follows: risk score = (exp Gene1 * coef Gene1) + (exp Gene2 * coef Gene2) +…+ (exp Gene N * coef Gene N).

All LNETs patients were divided into the high-risk and low-risk groups according to the median value of the risk score calculated from the formula. Overall survival (OS) differences between the high-risk and low-risk groups were assessed by Kaplan–Meier (K-M) method and two-tailed log-rank test. Time-dependent receiver operating characteristic (ROC) curve analysis within 1, 3 and 5 years to evaluate the predictive value of the prognostic risk model. The area under the ROC curve (AUC) ranges from 0.5 to 1, with 1 indicating perfect discrimination and 0.5 indicating no discrimination. Then, the prognostic risk model constructed by the training cohort was validated in the external dataset. Moreover, the risk model was validated in subgroups of carcinoids, LCNEC and SCLC.

Identification Of Predictive Factors Of Os

To identify independent prognostic predictors, univariate cox regression and subsequent multivariate cox regression were performed. The result was visualized by forest plot.

Gene Set Enrichment Analysis (Gsea)

Risk classification deriving from risk model was set as population phenotypes, and GSEA software (gsea.jar, version 3.0) was used to assess related pathways and molecular mechanisms in LNET patients. Enriched gene sets with nominal P value of < 0.05 and FDR of < 0.05 were considered statistically significant.

Statistical analysis

Statistical analyses were performed using R software (Version 3.6.3, R Foundation for Statistical Computing, Vienna, Austria). If not specified above, P < 0.05 was considered statistically significant.

Identification of DEGs and enrichment analysis in training cohort

The Kaplan-Meier survival analysis showed that the OS of carcinoids was significantly superior to that of LCNEC and SCLC (p < 0.001), while there was no statistically distinction between LCNEC and SCLC (p = 0.131) (Supplementary Fig. 1). 314 DEGs out of 20183 genes were screened out in the training cohort by using ‘edgeR’ package in R with P value < 0.05 and FDR < 0.05. For biological process, GO analysis indicated that those DEGs enriched mostly in (Fig. 1A). For cellular component and molecular function, the DEGs enriched mostly in (Fig. 1B) and (Fig. 1C). In addition, KEGG pathway analysis indicated that the DEGs were mainly enriched in (Fig. 1D).

Construction Of Prognostic Risk Model In Training Cohort

The univariate cox regression analysis of 314 DEGs in 40 high-grade LNET patients of training cohort screened out 75 prognosis-related DEGs. Subsequently, we used the least absolute shrinkage and selection operator (LASSO) method to further filter the prognosis-related DEGs, resulting in a model with 8 genes (SLC7A5, MEST, RAD51AP1, PMAIP1, GPER1, OIP5, IGLL3P, CDCA7) (Fig. 2A, 2B). Finally, a prognostic risk model was built as follows: risk score = SLC7A5 * (0.0975810777298566) + MEST * (0.141016923410625) + RAD51AP1 * (0.0738289697839021) + PMAIP1 * (0.00581195941306771) + GPER1 * (-0.24049872761534) + OIP5 * (0.0729507404044556) + IGLL3P * (-0.0979494191126777) + CDCA7 * (0.0117057735690952). The expression of SLC7A5, MEST, RAD51AP1, PMAIP1, OIP5, IGLL3P and CDCA7 were upregulated in tumor tissues, while the expression of GPER1 was downregulated in tumor tissues (Fig. 2C-J). The gene expression profiles in high-risk and low-risk groups were shown in a heatmap (Fig. 2K).

40 LNET patients were divided into high-risk and low-risk groups according to the median value of their risk score calculated from the prognostic risk formula. Kaplan-Meier survival analysis showed that OS in the high-risk group was significantly better than that in the low-risk group (p = 3e − 05) (Fig. 3A). To evaluate the predictive value of this prognostic risk model, time-dependent ROC curve analysis was performed and the AUC values of 1-, 3- and 5-year OS were 0.746, 0.926 and 0.917 (Fig. 3B-D).

Validation Of Prognostic Risk Model In Validation Cohort And Lnets Subgroups

The predictive value of this prognostic risk model was validated in the validation cohort composed of the other 40 high-grade LNETs patients. Similarly, by using the median risk score as the cutoff, 40 LNET patients were classified into the high-risk (n = 20) and low-risk groups (n = 20). K-M survival analysis revealed that patients in high-risk group had significantly higher mortality than patients with low risk group (p = 0.012) (Fig. 3E). Moreover, time-dependent ROC curve analysis revealed that the AUC values of 1-, 3- and 5-year OS were 0.758, 0.781 and 0.771 (Fig. 3F-H).

Since this prognostic risk model derived from the DEGs between high-grade LNETs and low-grade LNETs (carcinoids), we further check the predictive ability of the risk model in several subgroups of LNET patients. Based on the risk classification criteria and K-M survival analysis, a higher mortality was observed in LCNEC patients with high-risk, compared with those in low-risk group (p = 0.0015) (Fig. 4A), and the AUC values of 1-, 3- and 5-year OS were 0.715, 0.813 and 0.804 (Fig. 4B-D). Similar results were seen in 21 SCLC patients, with significant survival distinction (p = 0.013) (Fig. 4E) and AUC values of 0.769, 0.926 and 0.974 corresponding to 1-, 3- and 5-year OS (Fig. 4F-H). Interestingly, when this prognostic risk model was applied to carcinoid patient group, which was initially set as control group, a significant worse OS in high-risk carcinoid patients was observed (p = 0.003) (Fig. 4I), with AUC values of 0.957, 0.957 and 0.857 corresponding to 1-, 3- and 5-year OS (Fig. 4J-L). Moreover, when stratified by gene signature risk model, high-grade LNET patients with early stage (stage I and II) disease in high-risk group had significant worse survival (p = 0.047) (Supplementary Fig. 2A), while the gene signature stratified LNET patients with advanced stage (stage III and IV) disease into subgroups with significant distinction (p = 0.037) (Supplementary Fig. 2B).

Identification Of Prognostic Predictor

Due to the great performance of our gene signature risk model in predicting OS, we explored whether this risk model was an independent prognostic predictor in total 80 high-grade LNET patients. Clinicopathological features, including age, gender, TNM stage classification system and this risk model were enrolled in further prognostic analysis. As a result, univariate cox regression analysis showed that pathological stage (p = 0.002), N stage (p = 0.048), M stage (p = 0.024), as well as the risk model (p = 2.81E-05) were significantly associated with OS, while age (p = 0.24), gender (p = 0.16) and T stage (p = 0.067) were not (Table 1). Further multivariate cox regression analysis revealed that our gene signature risk model was the only prognostic predictor (p = 6.95E-04) (Table 1, Fig. 5).

Table 1

Univariate and multivariate cox regression analysis in high-grade LNET patients based on gene signature.
Variables	Univariate analysis			Multivariate analysis
Variables	HR	95% CI	P-value	HR	95% CI	P-value
age	1.36	0.81–2.28	0.238
gender	2.09	0.75–5.82	0.157
Pathological stage	2.28	1.35–3.85	0.002	2.23	0.76–6.59	0.146
T	1.63	0.97–2.77	0.068
N	1.73	1.00-2.98	0.048	0.62	0.22–1.74	0.365
M	3.94	1.19-13.00	0.024	1.75	0.48–6.41	0.395
riskScore	5.88	2.57–13.49	2.81E-05	4.66	1.92–11.35	6.95E-04

Gene Set Enrichment Analysis

GSEA was performed based on phenotypes of high risk and low-risk in the high-grade LNET patients. Genes in the high-risk group enriched mostly in the KEGG pathways of spliceosome, cell cycle, RNA polymerase, basal transcription factors, nucleotide excision repair and RNA degradation, responding the active tumor cell proliferation and metabolization (Fig. 6).

The survival of high-grade LNET patients was poor. The 5-year survival rate of LCNEC patients was about 20%(18, 19), while the overall 5-year survival rate of SCLC is a dismal 6%(20). Most high-grade LNET patients died within the first two years after diagnosis. Although the next generation sequencing was used to reveal the mRNA distinction between LNETs and normal tissues or between LNET subgroups, there was little progression of molecular target therapy and little improvement of practical treatment. Therefore, the research of genomic changes occurring in LNETs is in urgent need. In this study, we constructed an eight-gene signature by using univariate cox regression analysis and Lasso Cox regression model analysis for the prediction of OS in LNET patients. By applying this gene signature to the training cohort (n = 40), a significant better survival was observed in high-risk group compared with low-risk group. Similarly, a clear distinction was also observed between low-risk and high‐risk LNET patients (n = 40). Interestingly, the similar prognostic distinction between high-risk and low-risk groups was shown in low-grade LNET (carcinoids) patients too. When further stratified by different stage, both high-grade LNET patients with early stage and advanced stage disease in high-risk group had worse survival. The predictive value of this gene signature was validated by ROC curve analysis, indicating the practical applicability of this gene signature in LNET patients. The LNET patients in high risk could be identified and receive more aggressive treatments and closer follow-up to detect the tumor progression. Moreover, multivariate cox regression analysis revealed that our gene signature was the only prognostic predictor in high-grade LNETs, while stage, T, N and M categories of TNM classification were not.

By performing GO analysis of biological process, 314 DEGs enriched mostly in nuclear division and chromosome segregation. The KEGG pathway analysis of 314 DEGs enriched mostly in nuclear division, organelle fission and chromosome segregation. GSEA analysis in high-grade LNETs revealed that genes in high-risk group enriched significantly in spliceosome, cell cycle, RNA polymerase, basal transcription factors, nucleotide excision repair and RNA degradation. All these three analyses revealed the high activity of tumor cell division, proliferation and metabolization, indicating the malignancy of high-grade tumors and the potential therapeutic relevance of Cyclin-dependent kinases inhibitors in LNETs(21).

Since our eight-gene signature can effectively screen out high-risk LNET patients with worse prognosis, it is necessary to research the function and mechanism of these eight genes, including SLC7A5, MEST, RAD51AP1, PMAIP1, GPER1, OIP5, IGLL3P and CDCA7. A majority of genes in this signature were found to be involved in carcinogenesis and tumor progression. L-type amino acid transporter 1 (LAT1), which is encoded by SLC7A5, mediate the uptake of branch-chained and aromatic amino acids through a sodium-independent mechanism. LAT1 was overexpressed in about half of high-grade LNETs, and the expression of LAT1 correlated significantly with Ki-67(22). LAT1 has been proved to play an important role in cellular proliferation in several tumors(23–25). LAT1 not only supports mTORC1 activity, it also reinforces MYC and EZH2 signaling in cancer cells(26). MEST encodes a member of the alpha/beta hydrolase superfamily, which plays a role in development. In breast cancer, enhanced MEST expression is significantly associated with tumor pathogenesis, acts as a key regulator of IL-6/JAK/STAT3/Twist-1 signal pathway-mediated tumor metastasis and could induce tumor metastatic potential through induction of the epithelial-mesenchymal transition transcription factor and activation of the STAT3 nuclear translocation(27). RAD51AP1 is an accessory protein to RAD51 and has been shown to participate in the homologous recombination DNA damage response pathway and act as a growth promoting signaling molecule and is possibly critical during early stages of neoplasia(28, 29). RAD51AP1 is overexpressed in both tissue and peripheral blood of lung cancer patients and could predict worse overall survival. The silence of RAD51AP1 could reduce cell proliferation of lung cancer and downregulate pro-metastatic markers and the pro-survival mTOR signaling pathway in vitro(30). Noxa, encoded by PMAIP1 gene, is a pro-apoptotic protein in the B-cell lymphoma 2 family proteins, which belongs to a subclass of BH3-only proteins. Noxa plays a crucial role in the interactions with several proteins in the apoptosis pathway, highlighting its importance in the pathogenesis and treatment of certain cancers(31). Sensitivity to BH3 mimetic drugs, ABT-737, has been shown to be highly determined by Noxa expression in SCLC(32). G protein-coupled estrogen receptor 1 (GPER1) contributes to antiproliferation and autophagy in several cancer cells. GPER1 overexpressing breast cancer cells exhibit a reduced cell cycle progression, enhanced autophagy and reduced ability for energy production(33). Overexpression of GPER1 in lung cancer and may inhibit the hyperproliferation of cancer cells, with activation of this receptor leading to antiproliferative or proapoptotic effects(34). The protein encoded by OIP5 gene localizes to centromeres, where it is essential for recruitment of centromere Protein A through the mediator Holliday junction recognition protein. In several cancer types including lung cancer, OIP5 was overexpressed and the downregulation or silencing of OIP5 notably depresses cell survival and the colony-forming ability, while restoration of OIP5 expression stimulates cell growth(35–37). IGLL3P is a pseudogene lacking function annotations and remains to be clarified in future. CDCA7 is a c-Myc-responsive gene encoding cell division cycle-associated protein, which participates in cancer transformation and tumorigenesis in various types of human cancer(38–41). CDCA7 expression was upregulated in colorectal cancer tissues and the intensities of CDCA7 immunostaining were significantly associated with tumor invasion depth, lymph node metastasis and distant metastasis(41). Silencing CDCA7 inhibited cell proliferation in lung adenocarcinoma through G1 phase arrest and induction of apoptosis and CDCA7 could be used as a potential therapeutic target for new biomarkers(42).

Although the gene signature is promising, there are some limitations to our risk model. Firstly, due to the tumor heterogeneity and the mixed expression of tumor cells and normal cells(43), the predictive ability of our signature might not be stable. Secondly, because of the lacking data of LNET genomic information, we divided the GSE30219 dataset into training cohort and validation cohort at random by ‘sample’ function of R software, where contingency exists. Finally, considering the small number of LNET patients in our research, the gene signature risk model needs to be validated in large-scale clinical trials and prospective studies.

To summarize, based on the gene expression profile from GSE30219 dataset, we identified a novel eight-gene signature, which was promising in predicting survival in LNET patients. This gene signature might provide potential therapy target and relevant drugs. Moreover, this gene signature was the only predictor of survival and might help facilitate more personized treatment of LNET patients in practice.

Ethics approval and consent to participate The data were obtained from GSE30219 dataset, approval for our study by the ethics committee was not necessary.

Consent for publication All authors have seen the manuscript and approved to submit to your journal.

Availability of data and material The RNA sequencing data and clinical information of LNETs patients were extracted from GSE30219 dataset, which was downloaded from Gene Expression Omnibus (GEO) (http://www.ncbi.nlm.nih.gov/geo).

Competing interests The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Funding This work was supported by National High Level Hospital Clinical Research Funding (2022-PUMCH-B-011).

Authors' contributions Shanqing Li designed and directed this study. Guige Wang contributed to data collection, preprocess and visualization. Chao Guo analyzed and interpreted the results. Guige Wang and Chao Guo drafted and revised the manuscript. All authors have viewed and approved the final manuscript version.

Acknowledgement

We thank Rousseaux S et al. for their great contribution to the construction of GSE30219 dataset.

Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin. 2020;70:7–30.
Bray F, Ferlay J, Soerjomataram I, Siegel RL, Torre LA, Jemal A. Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin. 2018;68:394–424.
Oberg K, Modlin IM, De Herder W, Pavel M, Klimstra D, Frilling A, et al. Consensus on biomarkers for neuroendocrine tumour disease. Lancet Oncol. 2015;16:e435–46.
Travis WD, Brambilla E, Nicholson AG, Yatabe Y, Austin JHM, Beasley MB, et al. The 2015 World Health Organization Classification of Lung Tumors: Impact of Genetic, Clinical and Radiologic Advances Since the 2004 Classification. J Thorac Oncol. 2015;10:1243–60.
Fitzmaurice C, Akinyemiju TF, Al Lami FH, Alam T, Alizadeh-Navaei R, Allen C, et al. Global, Regional, and National Cancer Incidence, Mortality, Years of Life Lost, Years Lived With Disability, and Disability-Adjusted Life-Years for 29 Cancer Groups, 1990 to 2016: A Systematic Analysis for the Global Burden of Disease Study. JAMA Oncol. 2018;4:1553–68.
Asamura H, Kameya T, Matsuno Y, Noguchi M, Tada H, Ishikawa Y, et al. Neuroendocrine neoplasms of the lung: a prognostic spectrum. J Clin Oncol. 2006;24:70–6.
Derks JL, Hendriks LE, Buikhuisen WA, Groen HJ, Thunnissen E, van Suylen RJ, et al. Clinical features of large cell neuroendocrine carcinoma: a population-based overview. Eur Respir J. 2016;47:615–24.
Zaffaroni N, De Polo D, Villa R, Della Porta C, Collini P, Fabbri A, et al. Differential expression of telomerase activity in neuroendocrine lung tumours: correlation with gene product immunophenotyping. J Pathol. 2003;201:127–33.
Gridelli C, Rossi A, Airoma G, Bianco R, Costanzo R, Daniele B, et al. Treatment of pulmonary neuroendocrine tumours: state of the art and future developments. Cancer Treat Rev. 2013;39:466–72.
Simbolo M, Mafficini A, Sikora KO, Fassan M, Barbi S, Corbo V, et al. Lung neuroendocrine tumours: deep sequencing of the four World Health Organization histotypes reveals chromatin-remodelling genes as major players and a prognostic role for TERT, RB1, MEN1 and KMT2D. J Pathol. 2017;241:488–500.
Shimizu N, Akashi Y, Fujii T, Shiono H, Yane K, Kitahara T, et al. Use of ALK Immunohistochemistry for Optimal Therapeutic Strategy of Pulmonary Large-cell Neuroendocrine Carcinoma and Identification of a Novel KIF5B-ALK Fusion Oncokinase. Anticancer Res. 2019;39:413–20.
Swarts DR, Scarpa A, Corbo V, Van Criekinge W, van Engeland M, Gatti G, et al. MEN1 gene mutation and reduced expression are associated with poor prognosis in pulmonary carcinoids. J Clin Endocrinol Metab. 2014;99:E374-8.
Rindi G, Klersy C, Inzani F, Fellegara G, Ampollini L, Ardizzoni A, et al. Grading the neuroendocrine tumors of the lung: an evidence-based proposal. Endocr Relat Cancer. 2014;21:1–16.
George J, Lim JS, Jang SJ, Cun Y, Ozretić L, Kong G, et al. Comprehensive genomic profiles of small cell lung cancer. Nature. 2015;524:47–53.
Usuda J, Ichinose S, Ishizumi T, Ohtani K, Inoue T, Saji H, et al. Klotho is a novel biomarker for good survival in resected large cell neuroendocrine carcinoma of the lung. Lung Cancer. 2011;72:355–9.
Pietanza MC, Byers LA, Minna JD, Rudin CM. Small cell lung cancer: will recent progress lead to improved outcomes? Clin Cancer Res. 2015;21:2244–55.
Tibshirani R. The lasso method for variable selection in the Cox model. Stat Med. 1997;16:385–95.
García-Yuste M, Matilla JM, Alvarez-Gago T, Duque JL, Heras F, Cerezal LJ, et al. Prognostic factors in neuroendocrine lung tumors: a Spanish Multicenter Study. Spanish Multicenter Study of Neuroendocrine Tumors of the Lung of the Spanish Society of Pneumonology and Thoracic Surgery (EMETNE-SEPAR). Ann Thorac Surg. 2000;70:258–63.
Paci M, Cavazza A, Annessi V, Putrino I, Ferrari G, De Franco S, et al. Large cell neuroendocrine carcinoma of the lung: a 10-year clinicopathologic retrospective study. Ann Thorac Surg. 2004;77:1163–7.
Gaspar LE, McNamara EJ, Gay EG, Putnam JB, Crawford J, Herbst RS, et al. Small-cell lung cancer: prognostic factors and changing treatment over 15 years. Clin Lung Cancer. 2012;13:115–22.
Chou J, Quigley DA, Robinson TM, Feng FY, Ashworth A. Transcription-Associated Cyclin-Dependent Kinases as Targets and Biomarkers for Cancer Therapy. Cancer Discov. 2020;10:351–70.
Kaira K, Oriuchi N, Imai H, Shimizu K, Yanagitani N, Sunaga N, et al. Expression of L-type amino acid transporter 1 (LAT1) in neuroendocrine tumors of the lung. Pathol Res Pract. 2008;204:553–61.
Marshall AD, van Geldermalsen M, Otte NJ, Anderson LA, Lum T, Vellozzi MA, et al. LAT1 is a putative therapeutic target in endometrioid endometrial carcinoma. Int J Cancer. 2016;139:2529–39.
Yazawa T, Shimizu K, Kaira K, Nagashima T, Ohtaki Y, Atsumi J, et al. Clinical significance of coexpression of L-type amino acid transporter 1 (LAT1) and ASC amino acid transporter 2 (ASCT2) in lung adenocarcinoma. Am J Transl Res. 2015;7:1126–39.
Xu M, Sakamoto S, Matsushima J, Kimura T, Ueda T, Mizokami A, et al. Up-Regulation of LAT1 during Antiandrogen Therapy Contributes to Progression in Prostate Cancer Cells. J Urol. 2016;195:1588–97.
Salisbury TB, Arthur S. The Regulation and Function of the L-Type Amino Acid Transporter 1 (LAT1) in Cancer. Int J Mol Sci, 2018; 19.
Kim MS, Lee HS, Kim YJ, Lee DY, Kang SG, Jin W. MEST induces Twist-1-mediated EMT through STAT3 activation in breast cancers. Cell Death Differ. 2019;26:2594–606.
Wiese C, Dray E, Groesser T, San Filippo J, Shi I, Collins DW, et al. Promotion of homologous recombination and genomic stability by RAD51AP1 via RAD51 recombinase enhancement. Mol Cell. 2007;28:482–90.
Pathania R, Ramachandran S, Mariappan G, Thakur P, Shi H, Choi JH, et al. Combined Inhibition of DNMT and HDAC Blocks the Tumorigenicity of Cancer Stem-like Cells and Attenuates Mammary Tumor Growth. Cancer Res. 2016;76:3224–35.
Chudasama D, Bo V, Hall M, Anikin V, Jeyaneethi J, Gregory J, et al. Identification of cancer biomarkers of prognostic value using specific gene regulatory networks (GRN): a novel role of RAD51AP1 for ovarian and lung cancers. Carcinogenesis. 2018;39:407–17.
Morsi RZ, Hage-Sleiman R, Kobeissy H, Dbaibo G. Noxa: Role in Cancer Pathogenesis and Treatment. Curr Cancer Drug Targets. 2018;18:914–28.
Nakajima W, Hicks MA, Tanaka N, Krystal GW, Harada H. Noxa determines localization and stability of MCL-1 and consequently ABT-737 sensitivity in small cell lung cancer. Cell Death Dis. 2014;5:e1052.
Wegner MS, Gruber L, Schömel N, Trautmann S, Brachtendorf S, Fuhrmann D, et al. GPER1 influences cellular homeostasis and cytostatic drug resistance via influencing long chain ceramide synthesis in breast cancer cells. Int J Biochem Cell Biol. 2019;112:95–106.
Jala VR, Radde BN, Haribabu B, Klinge CM. Enhanced expression of G-protein coupled estrogen receptor (GPER/GPR30) in lung cancer. BMC Cancer. 2012;12:624.
Koinuma J, Akiyama H, Fujita M, Hosokawa M, Tsuchiya E, Kondo S, et al. Characterization of an Opa interacting protein 5 involved in lung and esophageal carcinogenesis. Cancer Sci. 2012;103:577–86.
Gong M, Xu Y, Dong W, Guo G, Ni W, Wang Y, et al. Expression of Opa interacting protein 5 (OIP5) is associated with tumor stage and prognosis of clear cell renal cell carcinoma. Acta Histochem. 2013;115:810–5.
Zheng YQ, Cui YR, Yang S, Wang YP, Qiu YJ, Hu WL. Opa interacting protein 5 promotes metastasis of nasopharyngeal carcinoma cells by promoting EMT via modulation of JAK2/STAT3 signal. Eur Rev Med Pharmacol Sci. 2019;23:613–21.
Jiménez PR, Martín-Cortázar C, Kourani O, Chiodo Y, Cordoba R, Domínguez-Franjo MP, et al. CDCA7 is a critical mediator of lymphomagenesis that selectively regulates anchorage-independent growth. Haematologica. 2018;103:1669–78.
Prescott JE, Osthus RC, Lee LA, Lewis BC, Shim H, Barrett JF, et al. A novel c-Myc-responsive gene, JPO1, participates in neoplastic transformation. J Biol Chem. 2001;276:48276–84.
Cheng C, Zhou Y, Li H, Xiong T, Li S, Bi Y, et al. Whole-Genome Sequencing Reveals Diverse Models of Structural Variations in Esophageal Squamous Cell Carcinoma. Am J Hum Genet. 2016;98:256–74.
Li S, Huang J, Qin M, Zhang J, Liao C. High expression of CDCA7 predicts tumor progression and poor prognosis in human colorectal cancer. Mol Med Rep. 2020;22:57–66.
Wang H, Ye L, Xing Z, Li H, Lv T, Liu H, et al. CDCA7 promotes lung adenocarcinoma proliferation via regulating the cell cycle. Pathol Res Pract. 2019;215:152559.
Fox NS, Haider S, Harris AL, Boutros PC. Landscape of transcriptomic interactions between breast cancer and its microenvironment. Nat Commun. 2019;10:3116.

Supple.Fig.1.tif
Supplementary Figure 1. Kaplan-Meier survival analysis of different subgroups of 104 LNET patients from GSE30219 dataset. The OS of carcinoids was significantly superior to that of LCNEC and SCLC, while there was no statistically distinction between LCNEC and SCLC.
Supple.Fig.2.tif
Supplementary Figure 2. Kaplan-Meier survival analysis of early-stage and advanced-stage high-grade LNET patients. The overall survival of patients in high-risk group was significantly worse than that in low-risk group both in high-grade LNET patients with early stage disease (A) and advanced stage disease (B).

Download PDF

Editorial decision: Major revision
13 Jan, 2023
Reviewers agreed at journal
06 Jan, 2023
Reviewers invited by journal
06 Jan, 2023
Editor assigned by journal
06 Dec, 2022
First submitted to journal
04 Dec, 2022

You are reading this latest preprint version

Identification of a novel prognostic gene signature in high-grade lung neuroendocrine tumors

Status:

Version 1

Abstract

Background

Materials and Methods

Results

Conclusion

Figures

Introduction

Materials And Methods

Data source and preprocessing

Identification Of Differentially Expressed Genes (Degs) And Enrichment Analysis

Construction And Validation Of Prognostic Degs Risk Model

Identification Of Predictive Factors Of Os

Gene Set Enrichment Analysis (Gsea)

Statistical analysis

Results

Identification of DEGs and enrichment analysis in training cohort

Construction Of Prognostic Risk Model In Training Cohort

Validation Of Prognostic Risk Model In Validation Cohort And Lnets Subgroups

Identification Of Prognostic Predictor

Gene Set Enrichment Analysis

Discussion

Conclusions

Declarations

References

Supplementary Files

Status:

Version 1