Clinical implications of single cell sequencing for bladder cancer

Bladder cancer (BC) is the 10th most common cancer worldwide, with about 0.5 million reported new cases and about 0.2 million deaths per year. In this scoping review, we summarize the current evidence regarding the clinical implications of single-cell sequencing for bladder cancer based on PRISMA guidelines. We searched PubMed, CENTRAL, Embase, and supplemented with manual searches through the Scopus, and Web of Science for published studies until February 2023. We included original studies that used at least one single-cell technology to study bladder cancer. Forty-one publications were included in the review. Twenty-nine studies showed that this technology can identify cell subtypes in the tumor microenvironment that may predict prognosis or response to immune checkpoint inhibition therapy. Two studies were able to diagnose BC by identifying neoplastic cells through single-cell sequencing urine samples. The remaining studies were mainly a preclinical exploration of tumor microenvironment at single cell level. Single-cell sequencing technology can discriminate heterogeneity in bladder tumor cells and determine the key molecular properties that can lead to the discovery of novel perspectives on cancer management. This nascent tool can advance the early diagnosis, prognosis judgment, and targeted therapy of bladder cancer.


Introduction
Single-cell sequencing technologies were pioneered in the early 90s and gained widespread attention in 2013 when they were named "Nature Methods".These technologies are superior to traditional bulk sequencing methods due to their ability to map heterogenicity among individual cells of the same tissue.In short, these technologies isolate cells before administrating highly sensitive next-generation sequencing (NGS), which can sequence the very small amount of material available at the single-cell level.There are numerous techniques for isolating cells.The simplest method is separating cells one by one by micropipette under a microscope.The limiting dilution technique involves separating and suspending the cells in liquid before distributing the suspension in multi-well plates at the concentration of 0.5 cells per well.Microfluidic separation is a more advanced high throughput technique with relatively low cost.Single-nucleus separating is sometimes used instead of whole-cell separating to increase the durability of samples [1,2].
RNA is the most commonly sequenced molecule in these technologies (RNAseq), allowing researchers to map gene activity at the time of biopsy in each cell (transcriptional variability).In the field of oncology, this allows for the assessment of the tumor microenvironment (TME) at the single-cell level.Whole-exome sequencing (WES) maps the functional DNA and whole-genome sequencing (WGS) maps the entire DNA.Targeted NGS (tNGS) maps a specific panel of related genes that may be altered in single cells.T cell receptor sequencing (TCR-seq) maps the TCR-related RNA activity of single T cells [2][3][4].Recently, NGS has been combined with the detection of other molecules such as proteins to concomitantly map them on individual cells (multi-omics).Chromatin immunoprecipitation (CHIP-seq), REAP-seq (RNA and surface protein), CITE-seq, and QuRIE-seq (RNA and epitopes) are among these methods [5][6][7][8].
Bladder cancer (BC) accounts for 3% of all new cancer diagnoses with 0.5 million new cases and 0.2 million deaths per year [9].It is the 13th leading cause of cancer-related death in the world [10].The 5-year overall survival rate for BC can be as high as 77% in non-metastatic disease and only 6% in metastatic BC.This cancer is four times more common in men (9.6 per 100,000) than in women (2.4 per 100,000).Tobacco smoking, positive family history, and exposure to chemicals are the main risk factors for this malignancy [11].The frequency of BC rises with age; more than 90% of diagnoses occur in patients over 55 years of age [11].Bladder cancer is classified into muscle-invasive bladder cancer (MIBC) and non-muscle-invasive bladder cancer (NMIBC) [9,12].In addition to surgery, patients may receive radiotherapy, chemotherapy with mitomycin C, epirubicin, thiotepa, gemcitabine, platinum-based agents, doxorubicin, Bacillus Calmette Guerin (BCG) immunotherapy, and Immune Checkpoint Inhibitor (ICI) THERAPY [13].Despite advances in available therapeutic methods, the clinical outcome of patients with bladder cancer remains unsatisfactory.
In this review, we summarize the clinical implications of single-cell sequencing technologies in bladder cancer.

Methods
The design and implementation of this scoping review were guided by Arksey et al.'s [14] methodological framework, described by the Joanna Briggs Institute [15].To organize the information, the PRISMA model [16] was used and we followed the PRISMA-ScR recommendations [17].

Inclusion and exclusion criteria
This review focuses on the clinically significant application of single-cell sequencing technologies in bladder cancer.Inclusion criteria are in vitro or in vivo studies involving single-cell sequencing in bladder cancer published in any language.Exclusion criteria are letters, reviews, conference abstracts, editorials, unrelated articles, replies, unpublished articles, and articles without full-text access.

Information sources
A comprehensive search of the literature published until February 2023 was conducted using Emtree language and medical subject headings (MeSh), in Pubmed, Embase, and the Cochrane Central Register of Controlled Trials (CENTRAL) databases.A broad search strategy was performed using a combination of the following search terms: ((urothelial cancer OR bladder cancer) AND singlecell sequencing).We supplemented this search with manual searches of the relevant articles identified through the Scopus, and Web of Science.Details of the search strategy are available in Appendix 1.

Data extraction
We used the Rayyan platform to screen articles and extract data.The software automatically removed duplicates; the remaining exclusions were manually performed by two researchers.Titles and abstracts were separately screened by the first and third authors with a portion of articles double screened for application of inclusion and exclusion criteria.The second author separately screened all articles.Data were extracted by the first three authors.

Results and Discussion
Our initial search resulted in 247 citations (187 citations from electronic databases and 60 citations from other sources).After removing the duplicates, the first two authors independently reviewed the remaining 180 titles and extracted those that clearly did not meet the inclusion criteria.The first three authors independently reviewed the abstracts of 135 potentially eligible studies.The full text of the remaining 66 studies was obtained and evaluated in accordance with the inclusion criteria.Any disagreement over inclusion was resolved by discussion.Finally, 41 studies were included in the review (Fig. 1).
Public databases: 29 articles analyzed public cancer databases as part of their study.Gene Expression Omnibus (GEO) [59] is currently the largest database containing single-cell sequencing data of cancer patients.Other databases such as The Cancer Genome Atlas project (TCGA) [60], ArrayExpress [61], and Genotype Tissue Expression project (GTEx) [62] mainly contain bulk sequencing data.TISCH [63] and The Cancer Single-Cell State Atlas (CancerSEA) [64] are lesser-known public databases of single-cell RNA sequences focusing on TME.The website of CellPress provides a review of highlighted reveals by these pooled public databases [65].
The clinical implication of single-cell sequencing is evident in seven studies [18,19,25,32,42,55,56].These studies show that separating cells of a patient's tumor and characterizing cell subtypes can improve clinical decisionmaking by predicting the tumor's response to different treatments.

Limitations of studies
Current single-cell sequencing techniques are extremely complex and involve multiple steps.The process of separating cells could damage the cell contents.Two studies addressed this issue by performing single-nucleus separation instead of whole-cell separation; the nucleus is less prone to degradation during the separation procedure [18,20].
Analyzing the high-dimensional data generated by single-cell sequencing is extremely complex and suffers from "the curse of dimensionality" [67], i.e., too much information [2,3,[68][69][70].Most studies addressed this issue by filtering for genes of interest as the first step of their analysis.For example, Zhou et al. [42] gathered a list of senescence-related genes from the website CellAge (http:// genomics.senescence.info/cells)and only investigated related data in the TCGA database to generate their B cell senescence-related gene signature (Table 1).Other studies addressed this complexity through the reduction of dimensions (excluding insignificant genes).The most common dimension reduction techniques were clustering using the R package Seurat and visualizing using t-SNE [71].
Public databases such as TCGA and GEO pool data from various sequencing studies without providing sufficient data on heterogeneity among their techniques and patient factors.This is less important when sequencing DNA since DNA is a stable molecule designed by nature to maintain its  [32].
It is important to keep in mind that transcription in cells is highly dependent on environmental factors.Temperature, oxygen level, stress, and even the circadian rhythm (time of the day) can influence transcription.Many commonly used medications such as steroids directly affect transcription [73][74][75][76].As Yao et al. showed (Table 1), the over-thecounter drug acetaminophen influences the expression level of the gene, they found to be associated with BC tumor grade [52].

Conclusion
Single-cell analysis is a relatively new tool that aids in understanding the complexity of bladder cancer tumor's microenvironment.Included studies in our scoping review collectively show that identifying certain cell subtypes in these heterogenous tumors may help clinicians in predicting response to expensive or invasive therapies such as immune checkpoint inhibitors or surgery with or without neoadjuvant chemotherapy.However, there are currently no clinical trials showing the benefit of performing this test on the overall well-being of the patients.Future single-cell sequencing studies should include epidemiological data, environmental factors, as well as comprehensive clinical data to improve our understanding of bladder cancer etiology and its heterogenous response to treatment.

FIGURE 1 .
FIGURE 1. Process of identification and inclusion of studies-PRISMA flow diagram.

TABLE 1
The summary of the included articles