Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Sep 1.
Published in final edited form as: Urol Oncol. 2022 May 23;40(9):410.e19–410.e27. doi: 10.1016/j.urolonc.2022.04.013

Characterizing Molecular Subtypes of High-Risk Non-Muscle-Invasive Bladder Cancer in African American Patients

Sungyong You 1, Minhyung Kim 1, Steven Widen 2, Alexander Yu 3, Gloria C Galvan 1, Yunhee Choi-Kuaea 1, Eduardo J Eyzaguirre 4, Lars Dyrskjøt 5, David J McConkey 6, Woonyoung Choi 6, Dan Theodorescu 7, Keith S Chan 7, Yong Shan 3, Douglas S Tyler 8, Amanda M De Hoedt 9, Stephen J Freedland 1,9,10, Stephen B Williams 3
PMCID: PMC9741768  NIHMSID: NIHMS1852547  PMID: 35618577

Abstract

Background:

We sought to determine whether differences in subtype distribution and differentially expressed genes exist between African Americans (AAs) and European Americans (EAs) in patients with high-risk non-muscle-invasive bladder cancer (NMIBC).

Methods:

We performed a retrospective cohort study including 26 patients (14 AAs and 12 EAs) from the University of Texas Medical Branch and the Durham Veterans Affair Health Care System from 2010-2020 among treatment naïve, high-risk NMIBC. Profiled gene expressions were performed using the UROMOL classification system.

Results:

UROMOL racial subtype distributions were similar with class 2a being most common with 10 genes commonly upregulated in AAs compared to EAs including EFEMP1, S100A16, and MCL1 which are associated with progression to MIBC, mitomycin C resistance, and bacillus Calmette-Guerin (BCG) durability, respectively. We used single nuclei analysis to map the malignant cell heterogeneity in urothelial cancer which five distinct malignant epithelial subtypes whose presence have been associated with different therapeutic response prediction abilities. We mapped the expression of the 10 genes commonly upregulated by race as a function of the five malignant subtypes. This showed borderline (p = 0.056) difference among the subtypes suggesting AAs and EAs may be expected to have different therapeutic responses to treatments for bladder cancer. AAs were enriched with immune-related, inflammatory, and cellular regulation pathways compared to EAs, yet appeared to have reduced levels of the aggressive C3/CDH12 bladder tumor cell population.

Conclusions:

While premature, gene expression differed between AAs and EAs, supporting potential race-based etiologies for muscle-invasion, response to treatments, and transcriptome pathway regulations.

Keywords: Bladder Cancer, High-risk, Non-invasive, Subtypes, Race

INTRODUCTION

Bladder cancer (BC) is the 6th most common cancer and the 7th leading cause of cancer death in the US.1 The vast majority (~70%) of new BC cases are non-muscle invasive BC (NMIBC) that often have an indolent disease course. However, once BC invades into the muscle the 5-year risk of BC death is ~50%.2 NMIBC is a heterogeneous disease that can be stratified at the time of initial diagnosis based on findings from the transurethral resection of bladder tumor (TURBT) into low, intermediate and high-risk for progression to muscle invasion.3, 4 High-risk NMIBC may progress in 33% of patients at 10 years.5 Consequently, identifying those at the highest risk of progression to muscle-invasive BC (MIBC) is a major unmet clinical need. While risk stratification has used clinical determinants, several NMIBC classification systems have been proposed to reduce this heterogeneity through improved biological understanding of the disease.6–8 Prior studies examining transcriptomics in high-risk NMIBC patients have been limited by few patients progressing to MIBC.9, 10 Furthermore, these studies included limited numbers of African Americans (AAs), a demographic that has significantly worse overall and BC-specific survival outcomes.9–12 As a result, whether known subtypes apply to AAs also remains uncertain.

The updated 2021 UROMOL NMIBC classification system reported prognostic and potential treatment applications of RNA transcriptomic profiling of subtypes, independent of other molecular-omics, indicating an opportunity to determine whether unique biological gene expressions exist in AAs compared to European Americans (EAs).13 We, for the first time, describe subtype distribution and differential gene expressions (DEGs) of high-risk NMIBC according to race. We aim to determine if the UROMOL subtype distribution applies to high-risk NMIBC patients and if there exists a difference between AAs and EAs transcriptomic signatures based on differential RNA expression analysis. We hypothesized subtype classification systems will be applicable to both race groups, however, distribution of subtypes will differ between AA and EAs with high-risk subtypes predominating in AAs.

METHODS

Data Source

Patients with high-risk NMIBC were identified at the University of Texas Medical Branch (UTMB) Health system and the Durham Veterans Affair Health Care System (DVAHCS) from January 1, 2010 to December 31, 2020. The study was approved by the Institutional Review Board at UTMB and DVAHCS.

Study Cohort

To be included in the study cohort, patients needed to have a primary histology of transitional cell carcinoma at diagnosis, ≥ 18 years, with a diagnosis of high-risk NMIBC (as defined as cTa high grade and/or T1 and/or CIS) confirmed by a genitourinary pathologist, and have been treatment naive defined as never received immunotherapy, intravesical chemotherapy, and/or systemic chemotherapy prior to TURBT. Exclusion criteria included individuals < 18 years and children, individuals without BC, pregnant women, and vulnerable populations (including prisoners and incarcerated patients. The study cohort was limited by the number of available AAs diagnosed with high-risk NMIBC. Following chart review and screening of inclusion/exclusion criteria at each institution, 6 AAs were identified at UTMB, and 8 AAs were identified in the DVAHCS. These patients were stratified by gender (male/female), and we performed a greedy algorithm matching patient age between the AA and EA cohorts at each institution. The final study population consisted of a total of 26 patients (14 AAs and 12 EAs) matched on age and sex.

Baseline Characteristics

Chart review was performed to extract clinical information including patient age at initial diagnosis as defined by date of index TURBT, race, gender, and smoking status (Former, Current, Never). For clinical stage information we used the TNM classification system according to the 8th edition of the American Joint Committee on Cancer (AJCC) Cancer Staging Manual.14 Clinical stage for NMIBC patients was used to identify patients with high-risk NMIBC.12, 15 Clinical stage group for high-risk NMIBC were further subcategorized as presence vs. absence of CIS. The WHO 1973 BC grading system (1, 2, 3) was revised in 2004 (low and high).16–18 Thus, we used an accepted grade categorization as low (Grade 1 and low) and high (Grade 2, 3 and high) grade. 16–18

Disease Progression

Disease progression for both cohorts was defined as: 1) pathologic muscle invasive disease (i.e., T2 or greater) upstaged from index TURBT, 2) receiving definitive treatment as defined as radical cystectomy, radiation, or systemic chemotherapy, and/or 3) developing metastases from index TURBT.

RNA Extraction and Sequencing

RNA extraction was performed by macrodissection for all samples. RNA extraction, sequencing, and profiling were performed using formalin fixed paraffin embedded (FFPE) TURBT specimens at the Genomics Core at UTMB (UTMB cohort) and at Cedars Sinai (DVAHCS cohort), respectively. Sequencing results were subjected to quality assessment (Supplementary Figure 1) using MultiQC software.19 Sequence reads will be aligned to the GRCh38 build of the Homo sapiens genome using the STAR aligner.20 Only uniquely mapped reads were retained. Gene-wise counts were obtained using RSEM.21 Reads overlapping exons in annotation build 38.1 of NCBI RefSeq database were included. We performed a minimal pre-filtering to remove genes that had only 0 or 1 read for downstream analysis. Counts were converted to log2; trimmed mean of M values (TMM) was normalized using edgeR software.22 The tumor purity was estimated by using ESTIMATE software (https://bioinformatics.mdanderson.org/estimate/). Relative Purity Score (RPS) was computed using this formular as a proxy of tumor purity: Median estimate of tumor purity of UTMB and DVAHCS cohorts are around over 80% and 60%, respectively (Supplementary Figure 2).

UROMOL subtype distribution and Gene Expression Differences by race

Normalized profiled gene expressions of high-risk NMIBC by race were performed and applied to UROMOL classification using the UROMOL classifier.13 A differential gene expression analysis comparing AAs and EAs was performed by edgeR22 and genes with adjusted P-value <0.05 and absolute log2-fold-change > 1 were defined as differentially expressed genes (DEGs).

Transcriptomic Pathway Analysis

Racial differences in transcriptomic pathways were evaluated with gene set enrichment analysis (GSEA)23 using the molecular signature database hallmark gene set collection MSigDB.24 Enriched pathways in both cohorts were recorded using normalized gene expression for the genes and pathway activation Z-score for the pathways.25

Single Nuclei Analysis

Recently, we used single nuclei analysis to map the malignant cell heterogeneity in MIBC.26 This revealed five distinct malignant epithelial subtypes whose presence was associated with different therapeutic response prediction ability. Tumors showing enrichment of the recently described C3 (Cell-Cell-Communication, CDH12/N-Cadherin 2) subtype.26 defined patients with poor outcome following surgery. In contrast, C3-enriched tumors exhibited superior response to immune checkpoint therapy (ICT). We mapped the expression of the 10 genes commonly upregulated from both cohorts as a function of the five malignant subtypes to determine whether differences in AA vs EA gene expression of high-risk NMIBC were associated with features of MIBC subpopulations. To assess the differential expression of 10 genes in the 5 distinct cell subtypes (C3, KRT6A, Cycling, UPK, and KRT13), we first select individual cell population and displayed relative fraction of corresponding cell population using Seurat software (https://github.com/satijalab/seurat).

Statistical Analysis

Student’s t-test for continuous variables and Fisher’s exact test for categorical variables were used to evaluate baseline characteristics between races. All p-values reported were 2-sided with a p<0.05 considered statistically significant.

RESULTS

We observed similar subtype distribution by race (Table 1). The UROMOL subtype classification revealed a high prevalence of the class 2a subtype in both race groups (Table 1). Clinical progression was observed in the 2a subtype in the UTMB cohort, with 1 AA and 1 EA progressing to muscle-invasive BC (MIBC), and in the 2b subtype in the DVAHCS cohort, with 1 EA progressing to MIBC (Table 1). De-identified TMM normalized expression values and separation level values for each patient in both cohorts are included (Supplementary Figures 3, 4A, 4B, Supplementary Table 1).

Table 1:

Baseline patient characteristics

UTMB DVAHCS
Demographic information Total (N=26) AA (N=6) EA (N=6) P-value AA (N=8) EA (N=6) P-value
  Age at Diagnosis, mean (SD) 67.2 ± 9.18 65.7 ± 10.8 65.8 ± 10.1 0.979 68.9 ± 10.1 67.8 ± 7.0 0.833
  Gender 1.0 n/a
   Female 4 2 2 0 0
   Male 22 4 4 8 6
  Smoker 0.134 0.790
   Former 12 1 4 4 3
   Current 8 2 2 3 1
   Never 6 3 0 1 2
  Clinical grade* n/a n/a
   High 26 6 6 8 6
  Clinical stage group 1.0 0.429
   CIS +/− T1 or TaHG 15 1 1 8 5
   TaHG or T1, no CIS 11 5 5 0 1
  Progression 1.0 0.429
   No 23 5 5 8 5
   Yes 3 1 1 0 1
  Subtypes** 1.0 0.467
   1 1 0 1 0 0
   2a 14 4 4 4 2
   2b 7 1 0 4 2
   3 4 1 1 0 2
Open in a new tab

AA=African American; CIS=carcinoma in situ; DVAHS: Durham Veterans Affair Health Care System; EA=European American; NMIBC=non-muscle invasive bladder cancer; UTMB: University of Texas Medical Branch

*

The WHO 1973 BC grading system (1,2,3) was revised in 2004 (papillary urothelial neoplasia low malignant potential (PUNLMP), low and high grades). We used BC grade categorization as low (Grade 1, PUNLMP and low) and high (Grade 2, 3 and high) grade.

**

UROMOL subtype classification

We generated a heatmap of differential expressed genes between AAs and EAs in both cohorts (Figure 1). A total of 201 genes and 290 genes were discovered to be significantly differentially expressed in the UTMB and DVAHCS cohorts, respectively (Figure 2). Of these, only 10 genes were discovered to be commonly upregulated in AAs compared to EAs from both cohorts: MMP25, PLA2G4C, EFEMP1, HPCAL1, TNFRSF19, MCL1, MRPL9, SLC13A3, MT1X, S100A16 (Figure 2B). Gene set enrichment analysis (GSEA) of AAs compared to EAs with high-risk NMIBC using the molecular signature database hallmark gene set collection MSigDB demonstrated multiple enriched pathways in each cohort (Figure 3 and Supplementary Figure 5). AAs were enriched with immune-related, inflammatory, and cellular regulation pathways compared to EAs (Figure 3 and Supplementary Figure 5).

Figure 1.

Figure 1.

Open in a new tab

Heat map of differentially expressed genes between African Americans (AAs) and European Americans (EAs).

Figure 2.

Figure 2.

Open in a new tab

Differential expression of African Americans (AAs) versus European Americans (EAs) and 10 genes commonly up-regulated in AAs between the two cohorts.

Figure 3.

Figure 3.

Open in a new tab

Gene set enrichment analysis (GSEA) of African Americans (AAs) and European Americans (EAs) with high-risk NMIBC using the molecular signature database hallmark gene set collection MSigDB.

Mapping of the expression of the 10 genes commonly upregulated from both cohorts as a function of the five malignant subtypes showed borderline (p=0.056) differences among the subtypes, with AAs demonstrating reduced expression of genes associated with the recently described C3 (Cell-Cell-Communication, CDH12/N-Cadherin 2) subtype.26 (Supplementary Figure 6).

DISCUSSION

Disparities in BC care are well known with social determinants of health thought to explain differences in outcomes by race. Studies assessing biological differences by race are lacking. In this small study, we found UROMOL class 2a subtype to be most common among AAs and EAs which has been previously associated with worse progression-free survival compared to other subtypes.13 Importantly, we also found 10 commonly upregulated DEGs and differences in enriched pathways in AAs compared to EAs in both cohorts.

Our study has several important findings. Of the 10 commonly upregulated DEGs we found among AAs, 3 have been associated with NMIBC: EFEMP1, S100A16, and MCL1. EFEMP1 (fibulin-3) has been previously found to be overexpressed in patients with MIBC vs NMIBC by Han et al., 2017.27 This group further demonstrated in vitro that shRNA knockdown and transient overexpression of EFEMP1 resulted in an increasing dose-dependent pro-invasive and migratory activity in transitional cell carcinoma (TCC) lines without impacting cellular morphology or growth rate.27 Critically, they performed in vivo orthotopic injections into mice bladder of stably transduced shRNA EFEMP1 knockdown TCC lines (T24 and UMUC-13) and demonstrated decreased incidence of muscle invasion of these TCC lines compared to controls.27 These findings suggest a step-wise progression with increased EFEMP1 expression as a plausible mechanism contributing to risk of muscle invasion in AAs, in addition to a possible novel therapeutic target.

Wang et al. established a mitomycin C (MMC)-resistant cell line (M-RT4) and showed that upregulation of S100A16 proteins suppressed apoptosis via the AKT/Bcl-2 pathway.28 Furthermore, this group performed an in vitro siRNA knockdown study of S100A16 in these cell lines and reported significantly decreased half-maximal inhibitory concentration following MMC treatment via cell counting, suggesting restored MMC sensitivity following S100A16 suppression.28 Moreover, prior studies in other cancer types have shown significantly differential expression of S100A16 between AA and EA prostate cancer patients.29 In addition, the TNFRSF19 gene is in the Decipher test gene panel, which co-expression of Decipher test genes was strongly preserved by race.30 These findings support the hypothesis of biological race-based differences in NMIBC. These results may also explain disparate race-based outcomes secondary to in vivo MMC resistance of NMIBC as S100A16 gene expression was found to be upregulated in our AA NMIBC population. Further, S100A16 expression may be a candidate factor in generalized MMC sensitivity and MMC treatment.

Sanders and Frasier et al., 2021 sought to determine whether response to bacillus Calmette-Guérin (BCG) induction therapy was associated with mutational and/or expressional changes in treatment naive high-grade T1NMIBC.31 They performed a retrospective study, stratifying patients at their institution into non-durable responders as defined as patients with recurrence of urothelial carcinoma of the bladder (any stage or grade) during the two-year study and durable responders as defined as any recurrence of lower grade or stage than the index lesion, and subsequently performed targeted DNA sequencing and whole exome RNAseq analysis on patient high-grade T1 NMIBC prior to BCG induction.31 They found that predicted deleterious frameshift deletions of the N-terminal 1st exon believed to be involved in protein regulation of MCL1 were significantly and disproportionally higher in durable responders and was significantly associated with recurrence free survival following BCG induction.31 As MCL1 is an anti-apoptotic protein of the BCL2 family and its down-regulation is linked to tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) sensitization32, 33, others hypothesized that the loss of function mutations in MCL1 may result in sensitization to TRAIL mechanisms. This may be due to a key death ligand in granulocytes following induction of BCG and release of interferon signaling particles by cancer cells resulting in the anti-neoplastic properties of BCG induction.31 This mechanism may provide a basis by which BCG responsiveness could theoretically be enhanced.31 While mutational burden was not a focus of our study, our findings may support a dose-dependent explanation for disparate outcomes via MCL1 mediated mechanisms resulting in possible race dependent BCG induction non-durability as we report significantly increased transcriptomic expression in AAs compared to EAs. While preliminary in a small cohort, our findings are exciting as they offer potential novel mechanisms for NMIBC adaptive pretreatment immune resistance.34

We found 6 pathways that were commonly enriched in both cohorts using the molecular signature database hallmark gene set collection MSigDB: 1) TNF alpha Signaling via NF-kB, 2) IL6 JAK/STAT3 signaling, 3) IL2 STAT5 signaling, 4) G2M checkpoint, 5) E2F targets, and 6) Complement signaling. It is interesting to note that while TNF alpha signaling via granulocytes was hypothesized to play a role in apoptosis of BC as previously stated, enrichment of TNF alpha signaling via NF-kB, a well-known anti-apoptotic transcription regulator, in AAs compared to EAs suggests dual roles of TNF alpha with respect to both anti-apoptosis and apoptosis via different pathways. This hypothesis was further supported by the enrichment of the other pathways demonstrating mixed inflammatory, immune, and cellular progression signatures. Interestingly, the E2F target pathway in AAs was not previously described as one of the more enriched transcriptomic regulons in the four 2020 UROMOL subtypes; however, increased E2F3 gene expression was reported to be significantly correlated with class 2a tumors and copy number alterations (CNAs), suggesting a critical role of this pathway in genomic instability.13 Thus, the E2F pathway enrichment in AA transcriptomes may be associated with earlier genomic instability compared to EAs. Ultimately, our findings suggest AAs may be more responsive to therapies targeting these pathways compared to EAs.

We previously found the aggressive C3 population was associated with worse prognosis from surgical treatment and better prognosis with immune checkpoint therapy (ICT).26 Given that AAs appeared to have relatively reduced transcription levels of genes associated with the aggressive C3 bladder tumor cell population relative to EAs (supplementary figure 3), we speculated these patients may have better prognosis from surgical treatment but worse from ICT. These findings are hypothesis generating as they provide preliminary evidence towards potential application of MIBC intratumoral heterogeneity analysis in high-risk NMIBC to potentially predict approaches in treating advanced high-risk NMIBC.

Our findings must be interpreted within the context of the study design. First, we examined the distribution of subtypes by race in a relatively small cohort. Our findings are hypothesis generating and require further investigation in a larger cohort. However, the present study represents the largest bladder cancer subtype study by race which further highlights importance of investigation in other datasets. Second, we used the UROMOL classification system and further investigation using other classification systems may further elucidate our findings. Moreover, only 10 genes were differentially expressed in both cohorts which suggests significant heterogeneity between both cohorts which further impact of findings. Third, we attempted to validate our findings in other datasets (i.e. TCGA), however, limited numbers of AAs and further limited numbers of high-risk NMIBC to match by age, race, sex highlight the importance of investigating our findings in a larger sample within a racially diverse cohort such as the VA. The present data is therefore hypothesis generating and opens doors for future projects.

CONCLUSION

In summary, we found similar subtype distribution among high-risk NMIBC patients according to race. In one of the largest high-risk NMIBC AA cohorts reported to date, our investigation determined that gene expression differed by race, supporting potential novel race-based etiologies for differences in muscle-invasion, response to treatments, and transcriptome pathway regulations. Further biological studies in NMIBC molecular sub-stratification, associated treatment(s), and prognoses in a larger cohort are needed to support these hypotheses.

Supplementary Material

1

Supplementary Figure 1. RNAseq data quality metrics for UTMB and DVAHCS samples.

Supplementary Figure 2. Distribution of Relative Purity Scores for UTMB and DVAHCS tumor samples.

Supplementary Figure 3. TMM normalization of patient cohort transcriptome counts. EA = European American; AA = African American (above – UTMB cohort, below – DVAHCS cohort).

Supplementary Figure 4A. UTMB cohort NMIBC classification result based on 2020 UROMOL NMIBC classification and separation levels.

Supplementary Figure 4B. DVAHCS cohort NMIBC classification result based on 2020 UROMOL NMIBC classification and separation levels.

Supplementary Figure 5. The enrichment plots of the top 6 commonly enriched hallmark gene sets in the African Americans from both cohorts.

Supplementary Figure 6. Mapped expression of the 10 genes commonly up-DEGs in the African Americans (AAs) vs. European Americans (EAs) from both cohorts as a function of the five malignant subtypes.

Supplementary Table 1. High-risk NMIBC classification quantitative results based on the 2020 UROMOL NMIBC classification (A – UTMB, B – DVAHCS).

Highlights.

  • In a small cohort, UROMOL racial subtype distributions were similar with class 2a being most common with 10 genes commonly upregulated in African Americans (AAs) compared to European Americans (EAs) including EFEMP1, S100A16, and MCL1.

  • Expression of the 10 genes commonly upregulated by race as a function of the five malignant subtypes showed borderline (p = 0.056) difference among the subtypes suggesting AAs and EAs may be expected to have different therapeutic responses to treatments for bladder cancer.

  • These hypothesis generating results support further biological studies in a larger cohort to understand NMIBC molecular sub-stratification, associated treatment(s), and prognoses by race.

Acknowledgement:

This study was conducted with the support of a Department of Defense Peer Reviewed Cancer Research Program (PRCRP) Career Development Award (W81XWH1710576), The University of Texas Medical Branch Department of Surgery Seed Grant, and the National Institute of Health Loan Repayment Program (SBW); and partially by the National Institutes of Health (NIH) (5TL1TR001440-02) (AY). Opinions expressed in this manuscript are those of the authors and do not constitute official positions of the US Federal Government or the Department of Veterans Affairs.

Footnotes

Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Conflicts of Interest: The authors declare no potential conflicts of interest.

Data Sharing Statement:

The data sets generated and analyzed during the current study are not publicly available because they contain potentially identifying and sensitive information but are available from the corresponding author on reasonable request. Upon request from a qualified investigator, a limited data set will be created for that investigator’s use and shared pursuant to a Data Use Agreement (DUA) appropriately limiting use of the data set and prohibiting the recipient from identifying or re-identifying (or taking steps to identify or re-identify) any individual whose data are included in the data set. Investigators who request to use the data will be required to obtain institutional review board approval and sign the DUA before release of the data. Interested investigators are encouraged to directly contact the corresponding author.

REFERENCES

  • 1.Siegel RL, Miller KD, Jemal A: Cancer statistics, 2018. CA Cancer J Clin, 68: 7, 2018 [DOI] [PubMed] [Google Scholar]
  • 2.Clark PE, Agarwal N, Biagioli MC et al. : Bladder cancer. J Natl Compr Canc Netw, 11: 446, 2013 [DOI] [PubMed] [Google Scholar]
  • 3.Chang SS, Boorjian SA, Chou R et al. : Diagnosis and Treatment of Non-Muscle Invasive Bladder Cancer: AUA/SUO Guideline. J Urol, 196: 1021, 2016 [DOI] [PubMed] [Google Scholar]
  • 4.Babjuk M, Burger M, Zigeuner R et al. : EAU guidelines on non-muscle-invasive urothelial carcinoma of the bladder: update 2013. Eur Urol, 64: 639, 2013 [DOI] [PubMed] [Google Scholar]
  • 5.Chamie K, Litwin MS, Bassett JC et al. : Recurrence of high-risk bladder cancer: a population-based analysis. Cancer, 119: 3219, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Hedegaard J, Lamy P, Nordentoft I et al. : Comprehensive Transcriptional Analysis of Early-Stage Urothelial Carcinoma. Cancer Cell, 30: 27, 2016 [DOI] [PubMed] [Google Scholar]
  • 7.Hurst CD, Alder O, Platt FM et al. : Genomic Subtypes of Non-invasive Bladder Cancer with Distinct Metabolic Profile and Female Gender Bias in KDM6A Mutation Frequency. Cancer Cell, 32: 701, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Lotan Y, Boorjian SA, Zhang J et al. : Molecular Subtyping of Clinically Localized Urothelial Carcinoma Reveals Lower Rates of Pathological Upstaging at Radical Cystectomy Among Luminal Tumors. Eur Urol, 76: 200, 2019 [DOI] [PubMed] [Google Scholar]
  • 9.Mallin K, David KA, Carroll PR et al. : Transitional cell carcinoma of the bladder: racial and gender disparities in survival (1993 to 2002), stage and grade (1993 to 2007). J Urol, 185: 1631, 2011 [DOI] [PubMed] [Google Scholar]
  • 10.Williams SB, Huo J, Kosarek CD et al. : Population-based assessment of racial/ethnic differences in utilization of radical cystectomy for patients diagnosed with bladder cancer. Cancer Causes Control, 28: 755, 2017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Underwood W, Dunn RL, Williams C et al. : Gender and geographic influence on the racial disparity in bladder cancer mortality in the US. J Am Coll Surg, 202: 284, 2006 [DOI] [PubMed] [Google Scholar]
  • 12.Prout GR, Wesley MN, McCarron PG et al. : Survival experience of black patients and white patients with bladder carcinoma. Cancer, 100: 621, 2004 [DOI] [PubMed] [Google Scholar]
  • 13.Lindskrog SV, Prip F, Lamy P et al. : An integrated multi-omics analysis identifies prognostic molecular subtypes of non-muscle-invasive bladder cancer. Nat Commun, 12: 2301, 2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.Amin MB, Greene FL, Edge SB et al. : The Eighth Edition AJCC Cancer Staging Manual: Continuing to build a bridge from a population-based to a more “personalized” approach to cancer staging. CA Cancer J Clin, 67: 93, 2017 [DOI] [PubMed] [Google Scholar]
  • 15.Yee DS, Ishill NM, Lowrance WT et al. : Ethnic differences in bladder cancer survival. Urology, 78: 544, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Compérat E, Varinot J, Moroch J et al. : A practical guide to bladder cancer pathology. Nat Rev Urol, 15: 143, 2018 [DOI] [PubMed] [Google Scholar]
  • 17.Amin MB, McKenney JK, Paner GP et al. : ICUD-EAU International Consultation on Bladder Cancer 2012: Pathology. Eur Urol, 63: 16, 2013 [DOI] [PubMed] [Google Scholar]
  • 18.Amin MB, Trpkov K, Lopez-Beltran A et al. : Best practices recommendations in the application of immunohistochemistry in the bladder lesions: report from the International Society of Urologic Pathology consensus conference. Am J Surg Pathol, 38: e20, 2014 [DOI] [PubMed] [Google Scholar]
  • 19.Ewels P, Magnusson M, Lundin S et al. : MultiQC: summarize analysis results for multiple tools and samples in a single report. Bioinformatics, 32: 3047, 2016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Dobin A, Davis CA, Schlesinger F et al. : STAR: ultrafast universal RNAseq aligner. Bioinformatics, 29: 15, 2013 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Li B, Dewey CN: RSEM: accurate transcript quantification from RNA-Seq data with or without a reference genome. BMC Bioinformatics, 12: 323, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Robinson MD, McCarthy DJ, Smyth GK: edgeR: a Bioconductor package for differential expression analysis of digital gene expression data. Bioinformatics, 26: 139, 2010 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Subramanian A, Tamayo P, Mootha VK et al. : Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A, 102: 15545, 2005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Liberzon A, Subramanian A, Pinchback R et al. : Molecular signatures database (MSigDB) 3.0. Bioinformatics, 27: 1739, 2011 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Levine DM, Haynor DR, Castle JC et al. : Pathway and gene-set activation measurement from mRNA expression data: the tissue distribution of human pathways. Genome Biol, 7: R93, 2006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Gouin KH 3rd, Ing N, Plummer JT et al. : An N-Cadherin 2 expressing epithelial cell subpopulation predicts response to surgery, chemotherapy and immunotherapy in bladder cancer. Nat Commun, 12: 4906, 2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Han AL, Veeneman BA, El-Sawy L et al. : Fibulin-3 promotes muscle-invasive bladder cancer. Oncogene, 36: 5243, 2017 [DOI] [PubMed] [Google Scholar]
  • 28.Wang C, Zhu X, Li A et al. : S100A16 regulated by Snail promotes the chemoresistance of nonmuscle invasive bladder cancer through the AKT/Bcl-2 pathway. Cancer Manag Res, 11: 2449, 2019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Hardiman G, Savage SJ, Hazard ES et al. : A Systems Approach to Interrogate Gene Expression Patterns in African American Men Presenting with Clinically Localized Prostate Cancer. Cancers (Basel), 13, 2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Creed JH, Berglund AE, Rounbehler RJ et al. : Commercial Gene Expression Tests for Prostate Cancer Prognosis Provide Paradoxical Estimates of Race-Specific Risk. Cancer Epidemiol Biomarkers Prev, 29: 246, 2020 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Sanders JA, Frasier C, Matulay JT et al. : Genomic analysis of response to bacillus Calmette-Guerin (BCG) treatment in high-grade stage 1 bladder cancer patients. Transl Androl Urol, 10: 2998, 2021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Redelman-Sidi G, Glickman MS, Bochner BH: The mechanism of action of BCG therapy for bladder cancer--a current perspective. Nat Rev Urol, 11: 153, 2014 [DOI] [PubMed] [Google Scholar]
  • 33.Kim SH, Ricci MS, El-Deiry WS: Mcl-1: a gateway to TRAIL sensitization. Cancer Res, 68: 2062, 2008 [DOI] [PubMed] [Google Scholar]
  • 34.Kates M, Matoso A, Choi W et al. : Adaptive Immune Resistance to Intravesical BCG in Non-Muscle Invasive Bladder Cancer: Implications for Prospective BCG-Unresponsive Trials. Clin Cancer Res, 26: 882, 2020 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

1

Supplementary Figure 1. RNAseq data quality metrics for UTMB and DVAHCS samples.

Supplementary Figure 2. Distribution of Relative Purity Scores for UTMB and DVAHCS tumor samples.

Supplementary Figure 3. TMM normalization of patient cohort transcriptome counts. EA = European American; AA = African American (above – UTMB cohort, below – DVAHCS cohort).

Supplementary Figure 4A. UTMB cohort NMIBC classification result based on 2020 UROMOL NMIBC classification and separation levels.

Supplementary Figure 4B. DVAHCS cohort NMIBC classification result based on 2020 UROMOL NMIBC classification and separation levels.

Supplementary Figure 5. The enrichment plots of the top 6 commonly enriched hallmark gene sets in the African Americans from both cohorts.

Supplementary Figure 6. Mapped expression of the 10 genes commonly up-DEGs in the African Americans (AAs) vs. European Americans (EAs) from both cohorts as a function of the five malignant subtypes.

Supplementary Table 1. High-risk NMIBC classification quantitative results based on the 2020 UROMOL NMIBC classification (A – UTMB, B – DVAHCS).

NIHMS1852547-supplement-1.docx (881.4KB, docx)

Data Availability Statement

The data sets generated and analyzed during the current study are not publicly available because they contain potentially identifying and sensitive information but are available from the corresponding author on reasonable request. Upon request from a qualified investigator, a limited data set will be created for that investigator’s use and shared pursuant to a Data Use Agreement (DUA) appropriately limiting use of the data set and prohibiting the recipient from identifying or re-identifying (or taking steps to identify or re-identify) any individual whose data are included in the data set. Investigators who request to use the data will be required to obtain institutional review board approval and sign the DUA before release of the data. Interested investigators are encouraged to directly contact the corresponding author.

RESOURCES