High Frequency Mutant Genes in Urothelial Carcinoma Based On Genomic Landscape

Objective The new generation of sequencing technology has been applied to the study of genomic genetic characteristics of urothelial carcinoma for 20 years. Researchers at home and abroad have done a lot of research work. Analyzing and summarizing the research results, we can clarify the genes with high-frequency mutations, which is of great signicance for the screening of biomarkers and molecular targets of urothelial carcinoma. Method We will adopt the PICOS analysis method of evidence-based medicine; follow the principles of systematic evaluation and meta-analysis; formulate literature retrieval keywords and retrieval strategies; determine the inclusion criteria; and statistically analyze the name, mutation frequency, quantity, and the total number of times in repeated reports of signicant mutant genes in the genomic landscape. A total of 6254 cases of urothelial carcinoma were sequenced in the 27 theses selected. Sequencing methods include whole genome sequencing, whole exome sequencing, and target exome sequencing. 27 genomic landscapes of urothelial carcinoma showed that the number of signicant mutant genes was 5-58, with an average of 26 reported in each paper. There were 273 genes with signicant mutations in urothelial carcinoma, 65.57% (179 / 273) of which were reported only once and 34.43% (94 / 273) were reported more than twice. The top 7 genes most frequently reported were TP53, PIK3CA, FGFR3, KDM6A, ARID1A , RB1 and STAG2. There were 273 genes with signicant mutations in the genome of urothelial carcinoma, and biomarkers may be selected from 94 genes with high-frequency of mutations.


Introduction
Urothelial cells stretch across the renal pelvis, ureter, bladder, and parts of the urethra. Urothelial carcinoma (UC) can be divided into Upper tract urothelial carcinoma (UTUC) and lower urinary tract epithelial cancer [1] . Lower urinary tract urothelial carcinoma is mainly bladder cancer (Bladder cancer, BC), the incidence of which is higher and is divided into non muscle invasive bladder cancer(Non-muscle invasive bladder cancer, NMIBC) and muscle invasive bladder cancer (Muscle invasive bladder cancer, MIBC). In the past 20 years, the new generation of sequencing (NGS) has been used to analyze a large number of genomic genetic characteristics of the UC, and many mutations which are associated with the occurrence and development of tumor, tumor biological characteristics, and clinical prognosis have been found [2,3] . Many important research reports have described genomic landscape (GL) in detail. However, we found signi cant differences in the types and frequencies of mutated genes reported in these literatures. Meanwhile, we also found that the genomic mutation landscape of UTUC, BC, NMIBC and MIBC also had their own signi cant characteristics. Therefore, using the method of evidence-based medicine Picos analysis [4] , this paper systematically evaluated and meta-analyzed the literature on UC genomic detection in recent 10 years. Through the statistical analysis of the number and mutation frequency of signi cantly mutated genes displayed by GL in the literature, we attempted to obtain the key genes and mutation rules of common mutations in different histopathological types of UC, which will provide a theoretical basis for the screening of UC biomarkers and the study of molecular imaging targets.

Basic data included in the paper
A total of 5643 relevant literatures were retrieved and excluded one by one, of which 27 original literatures with GL and in uencing factors > 5 were included. A total of 6254 cases of urothelial carcinoma were studied in these 27 papers, including 919 cases of UTUC and 5335 cases of BC. In addition to NMIBC and MIBC, BC also included bladder inverted papilloma and urothelial papilloma, bladder adenocarcinoma, micropapillary urothelial carcinoma, and neuroendocrine BC. Metastatic urothelial carcinoma of the upper and lower urinary tracts was also included.
Sequencing methods include whole genome sequencing, whole exome sequencing, and target exome sequencing. In this paper, the genes with mutation frequency > 10% were identi ed as signi cant mutation genes. GL was drawn according to the signi cantly mutated genes. The map mainly showed the name, mutation frequency, mutation type of mutated genes, and their correlation with tumor clinical characteristics (Table 1).

Frequency of mutant gene reported
To further compare the genes in Table 1 one by one and calculate the frequency of repeated reports (   According to the analysis of

Report frequency and mutation frequency
According to the analysis of Table 2, the four most frequently reported genes in 27 papers were TP53, PIK3CA, FGFR3 and KDM6A. The mutation frequencies of the four signi cant mutated genes in each paper were shown in Table 3. According to the analysis in Table 2, the mutation frequency range and average mutation frequency of the six genes with the most frequent occurrence in the 27 papers were TP53 (17-83%, 43.7%), PIK3CA (6% -54%, 20.5%), FGFR3 (11% -97%, 27.4%) and KDM6A (9-52%, 28.3%) respectively, indicating that the mutation frequency of each signi cant mutant gene was quite different in different cohorts (the blank space in the table indicates that the gene mutation was not detected).
According to overall analysis, many research teams worldwide have systematically studied the mutation characteristics of UC genome gene mutations. All studies have used similar sequencing technology, but the detection rate and types of mutant genes were very different in different research cohorts. The reasons for these differences may be related to the heterogeneity of tumor tissue, pathogenic factors, and ethnic and regional differences. The accuracy and scienti city of sequencing, however need to be further veri ed.

Discussion
The Therefore, through literature retrieval, this paper obtained 27 valuable and high-level academic papers, following the PICOS method of evidence-based medicine for systematic evaluation and meta-analysis and trying to comprehensively understand the characteristics of gene mutations at the UC genome level so as to provide a theoretical basis for the screening of key molecules of UC and promote the clinical transformation of UC biomarkers, molecular imaging targets, and targeted therapeutic drugs.
In this study, 273 signi cantly mutated genes were reported in 27 papers. Unfortunately, 94 genes were reported twice or more in different literatures, accounting for only 34.43% of all signi cantly mutated genes, while 179 genes were reported only once in a single literature, accounting for 65.57%, indicating that the heterogeneity of UC genome mutation was very strong and the screening of UC key molecules may be limited to these 94 genes.
In terms of researchers and research objects, the research results mainly came from China, the United States, Germany, Britain, Spain, Canada, etc. It may be that the differences of race, region, and pathogenic factors will also attribute to the distinctions of UC gene mutation.
In addition, this paper studied the different histopathological types, including upper urinary tract urothelial carcinoma and bladder cancer. Bladder cancer includes NMIBC, MIBC, bladder inverted papilloma, urothelial papilloma, bladder adenocarcinoma, micro papillary urothelial carcinoma, neuroendocrine bladder cancer, and metastatic urothelial carcinoma of the upper urinary and lower urinary tract. Combined with the analysis of Table 3, the types and frequencies of tumor gene mutations in different histopathological types may also be distinct, and the analysis of different mutant gene combinations may be helpful for the diagnosis and molecular typing of different histopathological diseases. In 2016, Seiler et al. [32] established a clinicopathological classi er of KNN51 genome (51 genes) to predict positive lymph node metastasis. Kamoun et al. [33] established a MIBC classi cation system based on network analysis and divided MIBC into 6 molecular types. In fact, the detection of gene mutation alone cannot be used as a biomarker. The research method of protein genomics, the combination of genome and proteome, will facilitate to screen biomarkers.
In conclusion, although the 20-year study of urothelial carcinoma genomics has enabled us to have a profound interpretation of the molecular characteristics at the DNA level, we still lack a fundamental understanding of the heterogeneity and complexity of urothelial carcinoma genome.

Picos analysis method of evidence-based medicine
This study adopts the Picos method of evidence-based medicine. Picos stands for participants, intervention, control, outcome, and study design. The subjects (P) include urothelial carcinoma of upper urinary tract and bladder cancer; Intervention (I) means the method of genome sequence analysis involving whole genome, exon, and targeted sequencing technology; Control (C) is to compare and analyze the GL data of UTUC, BC, NMIBC, and MIBC; Outcome (O) includes gene name, mutation frequency, and other test data; Research design (S) mainly studies the original works with an impact factor greater than 5, excluding editorials, comments, and conference abstracts.

Thesis retrieval strategy
Based on the search keywords (Table 4), a systematic literature search of GeenMedical, PubMed, and Web of Science was conducted. We limited our searches to publication from January 2011 to January 2021.

Exclusion criteria
Following the principles of systematic evaluation and meta-analysis, all researchers participated in selecting papers. The exclusion criteria are original papers with similar research contents and impact factor < 5, excluding editorials, comments, and conference abstracts. The inclusion criteria are original papers with impact factor > 5 and including the standard detailed genome landscape map in the research results.