Identification of genes predicting chemoresistance and short survival in ovarian cancer

Introduction

Ovarian cancer (OC) is a lethal gynecological malignancy, with about 19,710 new cases and 13,270 deaths every year in United States (1). The standard first-line treatment of OC is surgery followed with platinum and taxane centered chemotherapy (2). Although these therapies can achieve complete remission at the initial stage, most of patients are likely to suffer tumor recurrence predominantly owning to the emergence of chemoresistance, which finally leads to poor prognosis (3). Thus, the factors contribute to the chemoresistance are often the reason for short survival of the OC patients.

Lots of factors are participated in the modulation of chemoresistance and thus many kinds of molecules can be the potential biomarkers for chemoresistance and survival, such as microRNAs, cell cycle and mitosis-molecules, cancer stem cell related molecules, the immune response related molecules and other cancer-associated molecules (4). However, poor sensitivity and lack of specificity are the limitation for majority of biomarkers that have been studied (2). Therefore, there is an ongoing need to identify factors that affect chemoresistance and survival in OC.

Open data and bioinformatics can boost advancements in basic science (5), and reuse of open data is very powerful (6). For example, based on 2,579 tumors from The Cancer Genome Atlas (TCGA) of four gynecological types plus breast, five molecular subtypes have been developed to assess the survival status of patients (7). On the basis of big data of TCGA and Gene Expression Omnibus (GEO), we previously discovered a group of genes relevant to chemoresistance and outcome in OC (8), and developed the multi-gene prognostic signatures in liver cancer (9).

In the present study, based on big data mining and large sample analysis, we identified hundreds of chemoresistant-related genes in OC. The role of those genes in prediction of chemoresistance and short survival was evaluated. We present this article in accordance with the REMARK reporting checklist (available at https://tcr.amegroups.com/article/view/10.21037/tcr-23-2157/rc).

Methods

Text mining

Coremine (http://www.coremine.com/medical/) was used for retrieval of genes potentially linked to chemoresistance in OC, by combination of “ovarian cancer” [‘Ovarian Neoplasms (alias Ovarian Cancer)’ (mesh); ‘Ovarian Neoplasms (alias Ovarian Cancer)’ (mesh); ‘Malignant neoplasm of ovary (alias Ovarian Cancer)’ (disease)] and “drug resistance” [‘drug resistance’ (mesh); ‘Drug Resistance, Neoplasm’ (mesh)] (P<0.05).

Data acquisition and large samples

TCGA ovarian cohort (10) of 489 OC patients with clinical data and gene expression was retrieved from cBioPortal (http://www.cbioportal.org/) (11,12), including “staging”, “grading”, “overall survival” and “primary treatment outcome”, in which a subgroup of 90 platinum resistant samples and 197 sensitive samples was included. Kaplan-Meier (KM) plotter (13) of 1,816 OC patients with survival data was retrieved for survival analysis, containing a subgroup of 1,656 patients of which the overall survival (OS) data were accessible, and 1,435 patients of whom the progression-free survival (PFS) data were available. The survival data of 1,816 OC patients with mRNA expressions were integrated from GEO (https://www.ncbi.nlm.nih.gov/geoprofiles/) (14,15) (GSE51373, GSE9891, GSE63885, GSE15622, GSE30161, GSE14764, GSE65986, GSE18520, GSE27651, GSE26712, GSE19829, GSE26193, GSE23554 and GSE3149) and the ovarian cohort of TCGA (10). The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Statistical analysis

Statistical analysis was performed with SPSS (v 20.0). Student’s t-test was used to determine the significant variance between the two sets of data. Survival analyses were carried out by the R survival package in R (version 3.3.1). The survival curves were evaluated by the KM method, which gene expression was divided into low and high expression according to the median value, in accordance with a previous study (16). The univariate survival analysis was applied to calculate the hazard ratio (HR) and 95% confidence intervals. The biological processes annotation and pathway enrichment were subsequently performed. Gene expression in prediction of chemoresistance was analyzed by receiver operating characteristic (ROC) plotter (http://www.rocplot.org/) (17), in which the transcriptomic data of 1,816 OC patients were included. The relapse status at six months was used as a cutoff for determination of patient’s response to therapy. And patients relapsed within six months were considered as non-responders to drugs.

Results

Retrieval of genes that potentially affect OC chemoresistance

A total of 1,319 genes which potentially affect OC chemoresistance were retrieved from Coremine database by use of keywords “ovarian cancer” and “drug resistance” (P<0.05). A total of 1,298 genes potentially linked to chemoresistance in OC were obtained from a previous study (18). The above two groups of genes were combined and a total of 2,431 genes were obtained after deletion of duplicates. The transcript expression data of the 2,431 genes with clinical factors in 489 OC samples (with a subgroup of 90 chemoresistant samples and 197 sensitive samples) were retrieved to analyze the relationships of these genes with chemoresistance and prognosis, and 2,218 genes of which the data were available were used for subsequent analyses.

Identification of chemoresistant-related genes that predict chemoresistance

There were 301 genes differentially expressed in chemoresistant samples interacting with each other

Function annotation and enrichment were performed. There were 353 biological processes significantly annotated from the 301 genes and proteins [false discovery rate (FDR) <0.01]. And among the top 17 processes (FDR <1.0e−06), at least 185 genes (61.5%) were responded to stimulus, 166 (55.2%) were involved in regulation of biological processes, 118 (39.2%) were development related, and 115 (38.2%) were responded to chemical processes (Figure 1A). These results at least partially provide the links between those biological processes with chemoresistance in OC, specifically for the genes in response to stimulus.

Figure 1 The 301 genes differentially expressed in 90 chemoresistant samples in contrast to 197 sensitive samples according to TCGA ovarian cancer cohort. (A) Function annotation and enrichment analyses revealed the top 17 biological processes (FDR <1×10⁻⁶) that were differentially expressed, bubbles in the same cluster are represented by the same color; (B) enriched top 10 typical pathways that correlated with ovarian cancer chemoresistance. The X-axis represents the number of genes. TCGA, The Cancer Genome Atlas; FDR, false discovery rate.

Pathway-based identification of 26 novel and key genes which contributed to chemoresistance

Fifty-three pathways were significantly enriched from the 301 genes (FDR <0.01), and among the top 16 pathways (FDR <0.001), 10 of them were typical pathways playing critical roles in OC resistance, which including PI3K-Akt signaling, apoptosis, platinum chemoresistance signaling, Ras signaling, mammalian target of rapamycin (mTOR) signaling, ErbB signaling, etc. (Figure 1B).

Gene distribution on the above 10 pathways related to chemoresistance in OC was comprehensively analyzed. Combination of all the genes in the 10 pathways indicated that 51 of the 301 genes were distributed in the 10 pathways. Further analysis based on previous studies was performed to reveal the associations of the 51 genes with cancer development, and the results indicated that the roles of 26 genes in OC chemoresistance were less reported. The transcriptional levels of the 26 genes in 90 chemoresistant OC samples and 197 sensitive samples are shown in Figure 2. Among these, 14 genes including MSH6, CASP2, EIF2AK3, SOS1, EIF2AK2, FLT1, FZD5, HSP90AA1, HSP90B1, KDR, PAX6, PCK1, SDC1 and WNT7A were significantly down-regulated in 90 resistant samples, and 12 genes including PARD6B, AKT1S1, CALML3, CSF3, GNG7, IHH, NRG1, PIK3CD, RIN1, RPS6KA1, STAT5A and TBP were significantly up-regulated.

Figure 2 Twenty-six genes significantly and differentially expressed in chemoresistant samples in contrast to the sensitive samples, in accordance with TCGA ovarian cancer cohort. Sensitive: 197 platinum sensitive ovarian cancer samples; resistant: 90 resistant samples. *, P<0.05; **, P<0.01. TCGA, The Cancer Genome Atlas.

Large sample-based identification of 13 genes that predicted chemoresistance

The roles of the above 26 genes in prediction of chemoresistance in OC were evaluated in a large sample of 1,347 OC patients. As shown in Figure 3, 13 of the 26 genes were identified to be potential predictive biomarkers of chemoresistance in OC. Consistent with their expressions in chemoresistant samples (Figure 2), low expression of SOS1, CASP2, MSH6, HSP90AA1, HSP90B1 and FLT1, and high expression of PARD6B, STAT5A, RPS6KA1, RIN1, PIK3CD, CALML3 and NRG1, could predict the emergence of chemoresistance (Figure 3A). In particular, on one hand, low expression of SOS1, CASP2 and MSH6, and high expression of PARD6B and STAT5A could be more excellent for predicting drug resistance to any drugs and platin [P<0.01, area under the curve (AUC) >0.6] (Figure 3B); while low expression of SOS1, MSH6, HSP90AA1 and HSP90B1, and high expression of STAT5A could be more excellent for prediction of taxane resistance (P<0.01, AUC >0.6) (Figure 3C). On the other hand, three genes including SOS1, MSH6 and STAT5A were excellent for predicting chemoresistance to any drugs, platin and taxane (P<0.01, AUC >0.6).

Figure 3 Role of genes in prediction of chemoresistance in ovarian cancer. Transcriptome-level data of 1,347 ovarian cancer patients were included according to ROC plotter. The relapse status at 6 months was used as a cut off for definition of patient’s response to therapy, and those relapsed within the 6 months were considered as non-responders. (A) Abnormal expression of genes predicts chemoresistance to any drugs (includes platin, taxane, docetaxel, paclitaxel, gemcitabine, topotecan and avastin), in which 130 non-responders and 1,217 responders were included. (B) Abnormal expression of genes predicts chemoresistance to platin, in which 114 non-responders and 1,095 responders were included. (C) Abnormal expression of genes predicts chemoresistance to taxane, in which 81 non-responders and 807 responders were included. AUC, area under the curve; FPR, false positive rate; TPR, true positive rate; ROC, receiver operating characteristic.

Identification of chemoresistant-related genes for predicting short survival

Forty-four genes which dysregulated in chemoresistant samples could predict short OS and disease-free survival (DFS)

The roles of the 2,218 genes in prediction of DFS and OS were determined in 489 OC samples of TCGA cohort. Of which, 207 genes were related to DFS (P<0.05), 249 genes were related to OS (P<0.05), and collectively a total of 380 genes were related to DFS and/or OS. Then, an intersection of these 380 genes with the 301 genes dysregulated in 90 chemoresistant samples were performed, and total of 84 common genes were identified.

Further analyses on these 84 genes were performed. On one hand, the expression of the gene in chemoresistant samples should match the poor status of prognosis. For example, a gene would be retained if it was highly expressed in 90 chemoresistant samples in contrast to 197 sensitive samples, and the high expression predicted a short survival. On the other hand, a gene would be retained if its association with prognosis in OC was poorly studied. After the selection, total of 44 genes which significantly dysregulated in chemoresistant samples and predicted short survival were identified (Table 1).

Among the 44 genes, 17 of them were prominently associated with both OS and DFS in 489 OC patients. Among them, low expression of 12 genes (CASP2, CHIT1, STAT1, MSH6, AADAC, CAPN13, GCH1, OR6F1, PHGDH, RNF148, SLAMF7, and WDR45B) which were down-regulated in resistant samples were correlated with poor prognosis, and high expression of five genes (LAYN, PARD6B, GDF6, LIPC and TENM3) which were up-regulated in resistant samples predicted poor prognosis (Tables 1,2, Figure 4). Ten genes were only significantly associated with OS, among those, low expression of six genes (ALDH5A1, TREML2, MRS2, TRIM27, CXCR4 and KCNE3) which were down-regulated in resistant samples predicted short OS, and high expression of four genes (SPOCK2, RPL23, TCF15 and ACSS3) which were up-regulated in resistant samples associated with short OS (Tables 1,3). Seventeen genes were only significantly relevant to DFS, among those, low expression of 10 genes (NCOA1, KCNJ16, C16ORF89, HIPK1, LARP4, LGR5, PSMD1, VTCN1, EIF2AK3 and SOS1) which were down-regulated in resistant samples predicted short DFS and high expression of seven genes (GRYAB, GAP43, ICAM5, CATSPERD, PSG1, RSL24D1 and SNHG29) which were up-regulated in chemoresistant samples predicted short DFS (Tables 1,4).

Figure 4 Kaplan-Meier analyses determined that seventeen genes were relevant to overall survival and disease-free survival in ovarian cancer, based on TCGA cohort of 489 patients. Gene expression was divided into low (L) and high (H) by the median value. DFS, disease-specific survival; OS, overall survival; TCGA, The Cancer Genome Atlas.

Large sample-based identification of 10 genes which significantly predicted short OS

The roles of the above 44 genes in prediction of prognosis were further verified in large samples of 1,816 OC patients, which included a subgroup of 1,656 specimens with OS data, and 1,435 specimens with PFS data. Firstly, among the 17 genes which were relevant to OS in TCGA cohort of 489 patients (Figure 4, Table 3), 10 of them (CASP2, CH1T1, SPOCK2, TREML2, RPL23, TCF15, ALDH5A1, MRS2, TRIM27 and STAT1) were consistently related to OS in 1,656 OC samples (Figure 5A), of which, four genes including CASP2, CHIT1, SPOCK2 and TREML2 were also significantly associated with PFS (Figure 5A). The relationships of the 17 genes associated with DFS in the 489 patients (Table 4) were also submitted to the analysis, and five of them including CRYAB, GAP43, NCOA1, ICAM5 and KCNJ16 were consistently associated with prognosis in the 1,816 patients, and the former three were associated with both OS and PFS (Figure 5B).

Figure 5 Genes correlated with OS and PFS in 1,816 ovarian cancer patients based on the KM plotters collection. Gene expression was divided into low (L) and high (H) by the median value. (A) Genes related to OS in 489 patients of TCGA cohort consistently predicted OS and/or PFS in 1,816 patients of KM plotter collection. (B) Genes related to DFS in 489 patients was also relevant to OS and/or PFS in 1,816 patients. OS, overall survival; PFS, progression-free survival; KM, Kaplan-Meier; TCGA, The Cancer Genome Atlas.

Discussion

In this study, a total of 2,218 genes which potentially affected chemoresistance in OC were identified based on text mining, and 301 of them were significantly dysregulated in 90 chemoresistant OC samples in contrast to 197 sensitive samples. Analysis by biological processes annotation suggested those genes could potentially respond to stimulus (Figure 1A). It has been proved that response to stimulus is implicated in regulation of chemoresistance in cancers (19,20). Furthermore, pathway enrichment of the 301 genes was performed and 10 pathways such as PI3K-Akt signaling, apoptosis, and platinum chemoresistance signaling were significantly enriched (Figure 1B), which are typical pathways involved in modulation of chemoresistance in OC. The results strongly supported the relevance of those genes with chemoresistance. Intersection of all genes of the 10 pathways with the 301 genes identified 51 genes, which included some typical chemoresistant related genes in OC, such as AKT1, PIK3CA and MAPK1 (21). Among the 51 genes, the associations of 26 genes (Figure 2) with chemoresistance in OC have been rarely reported. The results above suggested that, probably via interactions with the typical 10 drug-resistant pathways in OC, the 26 genes might be new targets for management of OC, particularly in chemoresistant patients.

Because of the heterogeneity of OC, identifying predictive biomarkers are important for the selection of suitable treatments to improve patient survival (22). The use of gene expression signatures of key pathways that contribute to chemoresistance as predictive biomarkers is a reasonable approach (22). Several genes were previously identified to predict chemoresistance. For example, a low-RAS signature was shown to associate with sensitivity to the AKT inhibitor MK2206 (23), and BRCAness gene-expression in OC was associated with responsiveness to platinum-based chemotherapy (24,25). Despite of those findings, the biomarkers for prediction of chemoresistance are still less understood. In this study, the above identified 26 genes were distributed in 10 chemoresistant-related pathways, and based on the large sample analysis, 13 of them were potentially the predictive biomarkers for chemoresistance (Figure 3). Especially the low expression of SOS1, CASP2 and MSH6, and high expression of PARD6B and STAT5A were excellent for predicting chemoresistance to any drugs and platin; low expression of SOS1, MSH6, HSP90AA1 and HSP90B1, and high expression of STAT5A were excellent for predicting taxane resistance. The associations of SOS1, CASP2, PARD6B, STAT5A, HSP90AA1 and HSP90B1 with chemoresistance in OC were rarely known, although a research suggested that the silence of MSH6 in Saccharomyces cerevisiae could increase the strain resistance to cisplatin, doxorubicin, and carboplatin (26). This is consistent with our findings. However, the roles of those genes in prediction of chemoresistance in OC have not been reported so far.

Chemoresistance is a main factor that contributes to short survival in OC. Thus, further analyses of 2,218 genes in 489 OC patients were performed to identify the candidates that affected both chemoresistance and survival, and 44 genes were identified to be relevant to OS and/or DFS (Table 1), and their relationships with chemoresistance and survival were limited. The correlation of the above 44 genes with survival of OC patients was further verified in large cohort of up to 1,816 specimens, and 10 genes including CASP2, CH1T1, SPOCK2, TREML2, RPL23, TCF15, ALDH5A1, MRS2, TRIM27 and STAT1 were consistently related to OS (Figure 5). The results are consistent with the findings in previous studies. For example, low expression of CASP2 associated with poor OS in gastric carcinoma (27), and high expression of SPOCK2 predicted poor prognosis in OC (28).

Among all the genes identified in this study, CASP2 was the only one that the results were positive in all of the analyses. In the TCGA cohort of 489 OC patients, CASP2 was significantly down-regulated in 90 chemoresistant samples in contrast to 197 sensitive samples, and its low expression predicted short DFS and OS. In large cohort up to thousands of OC patients, CASP2 expression was consistently lower in chemoresistant samples and its low expression predicted chemoresistance to any drugs and platin, and short OS and PFS as well. CASP2 is one of CASPS, which often acts as intrinsic initiators of apoptosis (29,30). CASP2 plays important roles in apoptotic as well as nonapoptotic processes including apoptosis, cell cycle, autophagy, DNA repair, regulation of oxidant levels and lipid biosynthesis (31). The roles of CASP2 in cancer remain a matter of controversy, because the gene normally produces two mRNA splice variants CASP2L and CASP2S. CASP2L normally promotes apoptosis, while CASP2S normally inhibits apoptosis (32). In OC, CASP2 is the target of miR-383, which is overexpressed in OC cells and samples. High level of miR-383 and low expression of the CASP2 improve cell invasion, cell cycle progression and cell proliferation (33). These results are basically consistent with the findings in this study.

Taken together, 13 genes that affected chemoresistance and predicted chemoresistance in OC were identified, especially the six excellent candidate genes SOS1, CASP2, PARD6B, STAT5A, HSP90AA1 and HSP90B1. A total of 44 genes which potentially contributed to chemoresistance and related to prognosis were identified, especially 10 genes including CASP2, CH1T1, SPOCK2, TREML2, RPL23, TCF15, ALDH5A1, MRS2, TRIM27 and STAT1 were consistently related to overall survival in a group of 1,656 patients. Finally, it is noteworthy that CASP2 was the only gene that the results were positively and consistently in all analyses. The genes discovered in this study might be developed to be predictive markers, prognostic markers and therapeutic targets in the clinical management of OC. Studies have shown that, the caspase-2 regulatory mechanism can induce OC cell death (34), Han et al. found that casp-2S affects cellular apoptosis through its interaction with membrane-associated cytoskeletal Fodrin protein (35). Future work will investigate the function of CASP2.

OC cells interact with their surrounding microenvironment through a complex communication mechanism, impacting the tumor’s response to drugs. OC cells can change the composition of their surrounding microenvironment (immune cells, stromal cells and vascular endothelial cells, etc.) through various means, such as releasing cytokines, VEGF and other angiogenic factors (36), secreting exosomes (37), etc., thus forming a favorable environment that promotes tumor growth and invasion. The poorly metabolized tumor microenvironment affects clinical prognosis by forming a barrier to tumor-infiltrating immune cells (38).

The genes related to chemotherapy resistance in OC show some correlation with tumor immune infiltration, tumor mutation burden, and microsatellite instability. Some studies suggest that chemo-resistant cancer cells often have higher levels of tumor immune infiltration and tumor mutation burden (39-41). Additionally, microsatellite instability is also associated with chemotherapy resistance. An increase in mutation burden and microsatellite instability could potentially impact tumor growth, replication, and treatment response (42). Future work will explore the association between CASP2 and tumor immune invasion, tumor mutation burden, and microsatellite instability to provide reference for OC treatment strategies.

Conclusions

The identified genes specifically the CASP2 might be potentially used as predictive markers, prognostic markers and therapeutic targets in management of OC.

Acknowledgments

Funding: This work was supported by the National Natural Science Foundation of China (Nos. 81860458, 81903644, and 82260721), Guangxi Natural Science Foundation (Nos. 2024GXNSFDA010045, 2021GXNSFAA075002, and 2018GXNSFAA281227), Program Foundation of Key Laboratory of High-Incidence Tumor Prevention and Treatment, Ministry of Education (Nos. GKE-ZZ202148, and GKE-ZZ202017) and Advanced Innovation Teams and Xinghu Scholars Program of Guangxi Medical University, Innovation Project of Guangxi Graduate Education (No. YCSW2023237).

Footnote

Reporting Checklist: The authors have completed the REMARK reporting checklist. Available at https://tcr.amegroups.com/article/view/10.21037/tcr-23-2157/rc

Peer Review File: Available at https://tcr.amegroups.com/article/view/10.21037/tcr-23-2157/prf

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://tcr.amegroups.com/article/view/10.21037/tcr-23-2157/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013).

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Siegel RL, Miller KD, Wagle NS, et al. Cancer statistics, 2023. CA Cancer J Clin 2023;73:17-48. [Crossref] [PubMed]
2. Chandra A, Pius C, Nabeel M, et al. Ovarian cancer: Current status and strategies for improving therapeutic outcomes. Cancer Med 2019;8:7018-31. [Crossref] [PubMed]
3. Xiang Y, Chen YJ, Yan YB, et al. MiR-186 bidirectionally regulates cisplatin sensitivity of ovarian cancer cells via suppressing targets PIK3R3 and PTEN and upregulating APAF1 expression. J Cancer 2020;11:3446-53. [Crossref] [PubMed]
4. Davidson B. Biomarkers of drug resistance in ovarian cancer - an update. Expert Rev Mol Diagn 2019;19:469-76. [Crossref] [PubMed]
5. Mangul S, Martin LS, Langmead B, et al. How bioinformatics and open data can boost basic science in countries and universities with limited resources. Nat Biotechnol 2019;37:324-6. [Crossref] [PubMed]
6. Rung J, Brazma A. Reuse of public genome-wide gene expression data. Nat Rev Genet 2013;14:89-99. [Crossref] [PubMed]
7. Berger AC, Korkut A, Kanchi RS, et al. A Comprehensive Pan-Cancer Molecular Study of Gynecologic and Breast Cancers. Cancer Cell 2018;33:690-705.e9. [Crossref] [PubMed]
8. Yin F, Yi S, Wei L, et al. Microarray-based identification of genes associated with prognosis and drug resistance in ovarian cancer. J Cell Biochem 2019;120:6057-70. [Crossref] [PubMed]
9. Liu M, Liu X, Liu S, et al. Big Data-Based Identification of Multi-Gene Prognostic Signatures in Liver Cancer. Front Oncol 2020;10:847. [Crossref] [PubMed]
10. Integrated genomic analyses of ovarian carcinoma. Nature 2011;474:609-15. [Crossref] [PubMed]
11. Cerami E, Gao J, Dogrusoz U, et al. The cBio cancer genomics portal: an open platform for exploring multidimensional cancer genomics data. Cancer Discov 2012;2:401-4. [Crossref] [PubMed]
12. Gao J, Aksoy BA, Dogrusoz U, et al. Integrative analysis of complex cancer genomics and clinical profiles using the cBioPortal. Sci Signal 2013;6:pl1. [Crossref] [PubMed]
13. Gyorffy B, Lánczky A, Szállási Z. Implementing an online tool for genome-wide validation of survival-associated biomarkers in ovarian-cancer using microarray data from 1287 patients. Endocr Relat Cancer 2012;19:197-208. [Crossref] [PubMed]
14. Barrett T, Wilhite SE, Ledoux P, et al. NCBI GEO: archive for functional genomics data sets--update. Nucleic Acids Res 2013;41:D991-5. [Crossref] [PubMed]
15. Barrett T, Suzek TO, Troup DB, et al. NCBI GEO: mining millions of expression profiles--database and tools. Nucleic Acids Res 2005;33:D562-6. [Crossref] [PubMed]
16. Henderson MJ, Haber M, Porro A, et al. ABCC multidrug transporters in childhood neuroblastoma: clinical and biological effects independent of cytotoxic drug efflux. J Natl Cancer Inst 2011;103:1236-51. [Crossref] [PubMed]
17. Fekete JT, Ősz Á, Pete I, et al. Predictive biomarkers of platinum and taxane resistance using the transcriptomic data of 1816 ovarian cancer patients. Gynecol Oncol 2020;156:654-61. [Crossref] [PubMed]
18. Lloyd KL, Cree IA, Savage RS. Prediction of resistance to chemotherapy in ovarian cancer: a systematic review. BMC Cancer 2015;15:117. [Crossref] [PubMed]
19. Liu D, Wang L, Zhong R, et al. Parallel microfluidic networks for studying cellular response to chemical modulation. J Biotechnol 2007;131:286-92. [Crossref] [PubMed]
20. Lahiani MH, Eassa S, Parnell C, et al. Carbon nanotubes as carriers of Panax ginseng metabolites and enhancers of ginsenosides Rb1 and Rg1 anti-cancer activity. Nanotechnology 2017;28:015101. [Crossref] [PubMed]
21. Liu X, Gao Y, Lu Y, et al. Oncogenes associated with drug resistance in ovarian cancer. J Cancer Res Clin Oncol 2015;141:381-95. [Crossref] [PubMed]
22. Cheaib B, Auguste A, Leary A. The PI3K/Akt/mTOR pathway in ovarian cancer: therapeutic opportunities and challenges. Chin J Cancer 2015;34:4-16. [Crossref] [PubMed]
23. Loboda A, Nebozhyn M, Klinghoffer R, et al. A gene expression signature of RAS pathway dependence predicts response to PI3K and RAS pathway inhibitors and expands the population of RAS pathway activated tumors. BMC Med Genomics 2010;3:26. [Crossref] [PubMed]
24. Konstantinopoulos PA, Spentzos D, Karlan BY, et al. Gene expression profile of BRCAness that correlates with responsiveness to chemotherapy and with outcome in patients with epithelial ovarian cancer. J Clin Oncol 2010;28:3555-61. [Crossref] [PubMed]
25. Weberpals JI, Clark-Knowles KV, Vanderhyden BC. Sporadic epithelial ovarian cancer: clinical relevance of BRCA1 inhibition in the DNA damage and repair pathway. J Clin Oncol 2008;26:3259-67. [Crossref] [PubMed]
26. Durant ST, Morris MM, Illand M, et al. Dependence on RAD52 and RAD1 for anticancer drug resistance mediated by inactivation of mismatch repair genes. Curr Biol 1999;9:51-4. [Crossref] [PubMed]
27. Wang Z, Ni F, Yu F, et al. Prognostic significance of mRNA expression of CASPs in gastric cancer. Oncol Lett 2019;18:4535-54. [Crossref] [PubMed]
28. Lou W, Ding B, Zhong G, et al. Dysregulation of pseudogene/lncRNA-hsa-miR-363-3p-SPOCK2 pathway fuels stage progression of ovarian cancer. Aging (Albany NY) 2019;11:11416-39. [Crossref] [PubMed]
29. Kim YR, Kim KM, Yoo NJ, et al. Mutational analysis of CASP1, 2, 3, 4, 5, 6, 7, 8, 9, 10, and 14 genes in gastrointestinal stromal tumors. Hum Pathol 2009;40:868-71. [Crossref] [PubMed]
30. Kumar S. Mechanisms mediating caspase activation in cell death. Cell Death Differ 1999;6:1060-6. [Crossref] [PubMed]
31. Tiwari M, Sharma LK, Vanegas D, et al. A nonapoptotic role for CASP2/caspase 2: modulation of autophagy. Autophagy 2014;10:1054-70. [Crossref] [PubMed]
32. Solier S, Logette E, Desoche L, et al. Nonsense-mediated mRNA decay among human caspases: the caspase-2S putative protein is encoded by an extremely short-lived mRNA. Cell Death Differ 2005;12:687-9. [Crossref] [PubMed]
33. Liu J, Dou Y, Sheng M. Inhibition of microRNA-383 has tumor suppressive effect in human epithelial ovarian cancer through the action on caspase-2 gene. Biomed Pharmacother 2016;83:1286-94. [Crossref] [PubMed]
34. Yang CS, Matsuura K, Huang NJ, et al. Fatty acid synthase inhibition engages a novel caspase-2 regulatory mechanism to induce ovarian cancer cell death. Oncogene 2015;34:3264-72. [Crossref] [PubMed]
35. Han C, Zhao R, Kroger J, et al. Caspase-2 short isoform interacts with membrane-associated cytoskeleton proteins to inhibit apoptosis. PLoS One 2013;8:e67033. [Crossref] [PubMed]
36. Nunes SC. Tumor Microenvironment - Selective Pressures Boosting Cancer Progression. Adv Exp Med Biol 2020;1219:35-49. [Crossref] [PubMed]
37. Nakamura K, Sawada K, Kinose Y, et al. Exosomes Promote Ovarian Cancer Cell Invasion through Transfer of CD44 to Peritoneal Mesothelial Cells. Mol Cancer Res 2017;15:78-92. [Crossref] [PubMed]
38. Kao KC, Vilbois S, Tsai CH, et al. Metabolic communication in the tumour-immune microenvironment. Nat Cell Biol 2022;24:1574-83. [Crossref] [PubMed]
39. Fan S, Gao X, Qin Q, et al. Association between tumor mutation burden and immune infiltration in ovarian cancer. Int Immunopharmacol 2020;89:107126. [Crossref] [PubMed]
40. Gao C, Li H, Liu C, et al. Tumor Mutation Burden and Immune Invasion Characteristics in Triple Negative Breast Cancer: Genome High-Throughput Data Analysis. Front Immunol 2021;12:650491. [Crossref] [PubMed]
41. Wang ZM, Xu QR, Kaul D, et al. Significance of tumor mutation burden and immune infiltration in thymic epithelial tumors. Thorac Cancer 2021;12:1995-2006. [Crossref] [PubMed]
42. Chen J, Apizi A, Wang L, et al. TCGA database analysis of the tumor mutation burden and its clinical significance in colon cancer. J Gastrointest Oncol 2021;12:2244-59. [Crossref] [PubMed]