Background
Ovarian cancer is one of the most malignant tumors in the female reproductive system and ranks second only to cervical cancer in global incidence and mortality [
1]. Due to a lack of early symptoms and effective early screening diagnostic methods, most patients with OC are found in the late stage, and the 5-year survival rate is only 20–25% [
2]. The main treatment method is a combination of tumor cell ablation and chemotherapy drugs, such as paclitaxel and platinum-based drugs. Despite the development of diagnostic and treatment technology, the mortality rate has not improved significantly [
3]. Therefore, searching for new treatment methods plays an important role in improving the prognosis of patients with OC.
Immunotherapy is a kind of therapy that can enhance the autoimmune ability of patients to kill or eliminate cancer cells. Immunotherapy includes many methods, such as tumor vaccine [
4], immunocytotherapy, therapeutic antibody, small synthetic molecule inhibitors, immune checkpoint inhibitors, etc. Among them, immunocheckpoint inhibitors play a very important role in tumor treatment. Immunocheckpoint inhibitors have been used in melanoma [
5], non-small cell lung cancer [
6], Hodgkin’s lymphoma [
7] and many other tumors. In recent years, immunotherapy, as a new treatment of ovarian cancer, has gradually attracted people’s attention and achieved some results in the treatment of ovarian cancer. Especially the inhibitors for immunocheckpoint of PD1/PDL1. Unfortunately, the overall response rate of patients to these inhibitors is still low [
8].
The tumor mutation burden (TMB) refers to the total number of replacement and insertion/deletion (indel) mutations per basic group in the exon coding region of the assessed gene in the genome of a tumor cell. Driver gene mutations can lead to the occurrence of tumors, but a large number of somatic mutations produce neoantigens, which activate CD8+ cytotoxic T cells and exert an anti-tumor effect mediated by T cells. Thus, more neoantigens are produced as the number of genetic variations increases, and the more likely it is that the immune system will recognize them. TMB was originally intended as a biomarker for predictive efficacy in patients with advanced melanoma treated with ipilimumab or tremelimumab. Patients with melanoma and a high TMB level tend to have better efficacy against PD-1/PD-L1 checkpoint inhibitors than patients with a low TMB level [
9]. In recent years, treatment with PD-1/PD-L1 checkpoint inhibitors has developed rapidly, opening a new chapter in the treatment of advanced OC, but patients have shown low objective response rates [
10]. Therefore, finding suitable biomarkers to screen the dominant population and improve the efficacy of immunotherapy is the top priority of immunotherapy for OC.
In this study, we calculated the TMB in 397 patients with OC in the TCGA database. Then, we investigated the relationship between TMB, prognosis, and clinicopathological parameters, such as grade, FIGO stage, lymphatic metastasis, and vascular invasion in patients with OC. Finally, we investigated the gene expression and tumor infiltrating immune cells (TIICs) related to TMB. After a comprehensive analysis of the TMB of OC cases in the TCGA database, we determined that TMB plays an important role in the malignant progression and prognosis of OC. Thus, monitoring patient mutation load can be used to provide more accurate immunotherapy.
Materials and methods
TCGA data acquisition
We downloaded the OC genetic mutation data, transcriptome data, and clinical data from the TCGA database (/) [
11]. The genetic mutation data contained 37,248 mutated genes. The transcriptome data included 307 cases of OC. The clinical data included age, sex, grade, FIGO stage, lymph node metastasis, and vascular invasion. The gene microarray data and corresponding clinical information of verifying cohorts GSE9891 [
12] and GSE26193 [
13] was downloaded from GEO database (
https://www.ncbi.nlm.nih.gov/geo/, RRID:SCR_007303). The data were standardized, and R software (R Foundation for Statistical Computing, Vienna, Austria, RRID:SCR_003302) was used for all operations.
Calculation of TMB in patients with OC
TMB was defined as the number of somatic, coding, base replacement, and insert-deletion mutations per megabase of the genome examined using non-synonymous and code-shifting indels under a 5% detection limit. We used R software and the following formula to calculate the TMB of the patients with OC:
$$ {\text{TMB}} = {\text{Sn}} \times {{1,000,000} \mathord{\left/ {\vphantom {{1,000,000} {\text{n}}}} \right. \kern-0pt} {\text{n}}} $$
where Sn represents the absolute number of somatic mutations, and n represents the number of exon base coverage depth ≥ 100×) [
14].
Differential analysis and Functional enrichment analysis
Ovarian cancer data were divided into two groups according to median TMB value. Through the algorithm of limma package, the differentially expressed genes were calculated, and the genes with logFDR < 0.05 and lg|Fold change| (log|FC|) > 1 were selected as the significantly differentially expressed genes. In order to better understand the function of the selected differentially expressed genes, we use enrich GO in the clusterprofiler package of R to perform the GO function enrichment analysis and KEGG pathway enrichment analysis. The false discovery rate (FDR) was less than 0.01.
Identification and verification of hub TMB‐related signature
The expression data and survival data of the selected differential genes were combined, and univariate Cox proportional hazards regression (PHR) analysis was performed to obtain survival-related genes. The genes with the p values (p < 0.001) were fitted in a multivariate Cox PHR model establish an risk score model. Kaplan–Meier survival curve was drawn to evaluate the difference of overall survival rate between high and low risk groups (p < 0.05). The receiver operating characteristic (ROC) curve was calculated to assess the predictive power of the risk score model. Finally, the result was test in verifying cohorts GSE9891 and GSE26193.
Estimate of immune cell infiltration
CIBERSORT is a deconvolution algorithm that combines the labeled genomes of different immune cell subpopulations to calculate the proportion of 22 immune cells in tissues. The 22 types of immune cells included: 7 types of T cells (CD8+ T cells, naive CD4+ T cells, resting memory CD4+ T cells, activated memory CD4+ T cells, follicle-assisted T cells, regulatory T cells, and γδT cells), 3 types of B cells (naive B cells, memory B cells, and plasma cells) NK cells (resting NK cells and activated NK cells), and various myeloid cells (monocytes, M0 macrophages, M1 macrophages, M2 macrophages, resting dendritic cells, activated dendritic cells, resting mast cells, activated mast cells, eosinophils, and neutrophils). In this study, the CIBERSORT online platform (
http://cibersort.stanford.edu/) was used to complete the calculation, and each sample was assigned a
p value. Samples with a CIBERSORT output value of p < 0.05 were screened for further analysis [
15].
Identification of potential compounds
CMAP database stores up gene expression profile data of human cell lines including MCF7, ssMCF7, PC3, HL60 and SKMEL5 processed by 1309 bioactive small molecules. Differentially expressed genes based on TMB value were divided into up- and downregulated groups. The probe IDs of the two groups genes were uploaded to the connectivity map website (
https://portals.broadinstitute.org/cmap/), respectively, and then a permuted results were obtained.
Statistical analyses
SPSS 23.0 software (SPSS Inc., Chicago, IL, USA) was used for data recording and analysis, and the Kolmogorov–Smirnov test was used to determine whether variables obeyed a normal distribution. If the data were normally distributed, the mean ± standard deviation was calculated and the independent sample t-test was used to detect differences between groups. If a normal distribution was not observed, the median value was presented, and the non-parametric rank sum test was used to detect differences between the groups. Comparisons of classified data between the groups were analyzed by the Chi square test, and p < 0.05 was considered significant. The follow-up endpoint was overall survival (OS), which refers to the time from the date of the definite diagnosis of OC patients to death from any cause or the end of the final follow-up. The survival curve was plotted by Kaplan–Meier method, and the differences between the groups were assessed by the log-rank test. Cox proportional hazard model was used to evaluate the effect of clinical variables and TMB level on the OS of the patients with OC, and a p-value < 0.05 was considered significant.
Discussion
OC is a highly malignant tumor that seriously threatens a woman’s health. There are no typical clinical symptoms and signs in its early stage. Once symptoms appear, most cases are in an advanced stage, and the mortality rate is the highest among gynecological malignant tumors [
20]. Although cytoreductive surgery and platinum-based combination chemotherapy have improved the 5-year survival rate of patients with OC, there has been no substantial progress in clinical diagnosis and treatment of OC [
21]. Therefore, finding new treatments is crucial to improve the survival rate of patients with OC.
Gene mutations are changes in the molecular structure of genes caused by the replacement, addition, or deletion of DNA base pairs. According to the way genetic information changes, gene mutations can be divided into three types: same sense mutations, missense mutations, and nonsense mutations [
22]. Same sense mutations do not have an actual mutation effect, while missense and nonsense mutations in most cases affect the structure and function of proteins or enzymes, thereby changing the genetic information. In our study, the mutations in the patients with OC were mainly missense mutations. The distribution of mutation sites in the gene is different, most of which occur on some mutation hot spots [
23]. Therefore, it is of great significance for diagnosis and treatment of tumor-related diseases to search for these hot mutated genes by gene sequencing technology. In our research, we found that TP53 had a high mutation frequency in patients with OC.
TMB is an important biological marker reflecting the degree of tumor mutation. Alexandrov and Lawrence et al. found that the TMB among tumor samples was significantly different, which was at least 0.001/Mb and up to 400/Mb. The TMB of different patients is also significantly different even for the same type of tumor. Some studies have reported that the TMB as a biological marker has an important correlation with the therapeutic effect of cancer immunotherapy [
24]. The reason why TMB is a marker of immunotherapy stems from the biological mechanism of somatic mutation and the immune response. Somatic mutations of tumors include synonymous mutations and non-synonymous mutations. Non-synonymous mutations produce abnormal proteins by changing the amino acid sequence. However, the immunogenicity of abnormal proteins in tumors is the basis of the tumor immune response. If abnormal proteins are finally recognized by immune cells, they will become neoantigens, and subsequent immune responses can develop [
25]. That is to say, when the TMB of a tumor sample is high, the mutations that produce immunogenic neoantigens in the mutations also increase. It is easier for the immune system to recognize and remove tumor cells, and the survival rate of patients will be relatively improved. In our study, the OS of the patients with OC in the high TMB group was significantly higher than that in the low TMB group, which was consistent with previous assumptions. However, We were unable to validate our predictions in other OC datasets due to the lack of prognostic information. In addition, we found that there was a statistical correlation between TMB and FIGO stages, Grade or tumor residual size. Then, five genes (RBMS3, PLA2G5, CDH2, AMHR2 and ADAMTS8) were selected to establish TMBRS model based on univariate and multivariate Cox PHR. The ROC curve and validation data sets all revealed that the TMBRS model was reliable in predicting recurrence risk. However, further we need more clinical trials to verify the results.
TIICs are part of the tumor microenvironment that promote, regulate, and inhibit the development and growth of tumors. According to the interactions between the types and functions of immune cells, immune cells may play a variety of roles in the development of tumors [
26]. In our study, we used the CIBERSORT algorithm to calculate the proportion of 22 immune cells in OC. The patients with OC were divided into two groups according to the TMB naive B cells, memory B cells, resting memory CD4+ T cells, Tregs, monocytes, resting mast cells, and neutrophils were higher infiltrating in low‐TMB group, while activated memory CD4+ T cells, follicle-assisted T cells, M1 macrophages were higher infiltrating in high‐TMB group, which indirectly confirms the previous view that a high TMB of tumors can induce the immune response of the body and thus inhibit the growth of tumors.
Conclusions
In conclusion, our results suggest that TMB, as an important biomarker of tumor mutation, plays an important role in the prognosis and guiding immunotherapy of OC. By determining the TMB of patients with OC, clinicians can more accurately treat patients with immunotherapy, thereby improving their survival rate.
Open AccessThis article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit
http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (
http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.