Construction of Decision Trees Based on Gene Expression Omnibus Data to Classify Bladder Cancer and Its Subtypes.

BACKGROUND Bladder cancer is a malignant tumor of the genitourinary system. Different subtypes of bladder cancer have different treatment methods and prognoses. Therefore, identifying hub genes affecting other genes is of great significance for the treatment of bladder cancer. MATERIAL AND METHODS We obtained expression profiles from the GSE13507 and GSE77952 datasets from the Gene Expression Omnibus database. First, principal component analysis was used to identify the difference in gene expression in different types of tissues. Differential expression analysis was used to find the differentially expressed genes between normal and tumor tissues, and between tumors with and without muscle infiltration. Further, based on differentially expressed genes, we constructed 2 decision trees for differentiating between tumor and normal tissues, and between muscle-infiltrating and non-muscle-infiltrating tumor tissues. A receiver operating characteristic curve was used to evaluate the prediction effect of the decision trees. RESULTS FAM107A and C8orf4 showed significantly lower expression in bladder cancer tissues than in normal tissues. Regarding muscle infiltration, CTHRC1 showed lower expression and HMGCS2 showed higher expression in non-muscle-infiltrating samples than in those with muscle infiltration. We constructed 2 decision trees for differentiating between tumor and normal tissue, and between tissues with and without muscle infiltration. Both decision trees showed good prediction results. CONCLUSIONS These newly discovered hub genes will be helpful in understanding the occurrence and development of different subtypes of bladder cancer, and will provide new therapeutic targets and biomarkers for bladder cancer.

Assessment of Luminal and Basal Phenotypes in Bladder Cancer.

Genomic profiling studies have demonstrated that bladder cancer can be divided into two molecular subtypes referred to as luminal and basal with distinct clinical behaviors and sensitivities to frontline chemotherapy. We analyzed the mRNA expressions of signature luminal and basal genes in bladder cancer tumor samples from publicly available and MD Anderson Cancer Center cohorts. We developed a quantitative classifier referred to as basal to luminal transition (BLT) score which identified the molecular subtypes of bladder cancer with 80-94% sensitivity and 83-93% specificity. In order to facilitate molecular subtyping of bladder cancer in primary care centers, we analyzed the protein expressions of signature luminal (GATA3) and basal (KRT5/6) markers by immunohistochemistry, which identified molecular subtypes in over 80% of the cases. In conclusion, we provide a tool for assessment of molecular subtypes of bladder cancer in routine clinical practice.

Immunohistochemical assessment of basal and luminal markers in non-muscle invasive urothelial carcinoma of bladder.

The Cancer Genome Atlas project introduced genomic taxonomy of basal and luminal molecular subtypes in muscle invasive bladder cancer. Fewer studies have addressed the molecular classification in non-muscle invasive bladder cancer (NMIBC). Our aim is to assess the applicability of the proposed phenotypic classification for NMIBC. Three TMAs were constructed from 193 TURBT specimens of 60 bladder cancer patients treated at one of the authors' institutions (1998-2008). Follow-up data on recurrence, grade, or stage progression was obtained. Immunohistochemistry was performed using an automated Ventana System for markers indicative of luminal (GATA3, CK20, ER, Uroplakin II, and HER2/neu) and basal (CK5/6 and CD44) phenotype. Marker expression was evaluated by 3 urologic pathologists. Using unadjusted logistic regression, we found significant association between tumor recurrence at next biopsy and CD44 expression (OR = 2.51, P = 0.03), tumor recurrence at any subsequent biopsy and ER expression (OR = 0.24, P = 0.04), and tumor grade progression at any subsequent biopsy and HER2/neu expression (OR = 0.24, P = 0.04). After adjusting for pathologic stage, we found a significant association between CK5/6 expression and tumor stage progression at either next or any subsequent biopsy (OR = 0.94, P = 0.006; and OR = 0.97, P = 0.02, respectively). Our findings suggest that individual immunohistochemical markers may be of value as prognostic factors in NMIBC.

Patient-Derived Non-Muscular Invasive Bladder Cancer Xenografts of Main Molecular Subtypes of the Tumor for Anti-Pd-l1 Treatment Assessment.

BACKGROUND: Establishment of heterotopic patient-derived xenografts of primary and relapsed non-muscular invasive bladder cancer (NMIBC) to explore the biological property of PD-L1 signaling that may impact bladder tumor growth in humanized animals. METHODS: Tumor cells of luminal, basal, and p53 subtypes of primary and relapsed NMIBC were engrafted to irradiated (3.5 Gy) NOG/SCID female mice along with intraperitoneal transplantation of human lymphocytes (5 × 107 cells/mouse); a role of PD-L1 signaling pathway inhibition for bladder cancer growth was assessed in humanized animals that carried PD-L1-expressing main molecular subtypes of bladder carcinoma patient-derived xenografts (PDX) and provided with selective anti-PD-L1 treatment. We used two-tailed Student's t test to explore differences between main and control subgroups. Significance of intergroup comparison was measured with one-way ANOVA followed by the Tukey's or Newman-Keul's criterion. Survival curves were analyzed with the Gehan's criterion with the Yate's correction. The Spearman's correlation was used to assess the link between CD8+ expression and sPD-L1 serum level. Differences were considered statistically significant at p < 0.05. RESULTS: Heterotopic primary and relapsed luminal, basal, and p53 subtypes of NMIBC PDXs were established. More than 25% of counted tumor cells of all PDX specimens expressed PD-L1, so the tumors were ranged as PD-L1 positive. Anti-PD-L1 intervention increased survival of the animals that carried both primary and relapsed luminal noninvasive, muscular invasive, and relapsed luminal bladder cancer xenografts. There was significant retardation of tumor volume duplication time in aforementioned subgroups correlated with PD-L1 expression. Bad response of p53 mutant subtypes of NMIBC on specific anti-PD-L1 treatment may be associated with low CD8+ cells representation into the tumors tissue. CONCLUSIONS: Established PD-L1-positive NMIBC PDXs differently replied on anti-PD-L1 treatment due to both NMIBC molecular subtype and tumor T-suppressors population. The results may have major implications for further clinical investigations.

Evaluating the cost of surveillance for non-muscle-invasive bladder cancer: an analysis based on risk categories.

INTRODUCTION: Non-muscle-invasive bladder cancer (NMIBC) is a biologically heterogeneous disease and is one of the most expensive malignancies to treat on a per patient basis. In part, this high cost is attributed to the need for long-term surveillance. We sought to perform an economic analysis of surveillance strategies to elucidate cumulative costs for the management of NMIBC. METHODS: A Markov model was constructed to determine the average 5-year costs for the surveillance of patients with NMIBC. Patients were stratified into low, intermediate, and high-risk groups based on the EORTC risk calculator to determine recurrence and progression rates according to each category. The index patient was a compliant 65-year-old male. A total of four health states were utilized in the Markov model: no evidence of disease, recurrence, progression and cystectomy, and death. RESULTS: Cumulative costs of care over a 5-year period were $52,125 for low-risk, $146,250 for intermediate-risk, and $366,143 for high-risk NMIBC. The primary driver of cost was progression to muscle-invasive disease requiring definitive therapy, contributing to 81% and 92% of overall cost for intermediate- and high-risk disease. Although low-risk tumors have a high likelihood of 5-year recurrence, the overall cost contribution of recurrence was 8%, whereas disease progression accounted for 71%. CONCLUSION: Although protracted surveillance cystoscopy contributes to the expenditures associated with NMIBC, progression increases the overall cost of care across all three patient risk groups and most notably for intermediate- and high-risk disease patients.

Immunochemical and molecular assessment of urothelial neoplasms and aspects of the 2016 World Health Organization classification.

The new World Health Organization classification of tumours of the urinary system and male genital organs (4th edn) has several changes from previous versions, and was published in January 2016. New pathways have been discovered in the development of bladder cancer, and were included in this new classification. Guidance from the International Collaboration on Cancer Reporting (ICCR) helped to clarify open questions, in conjunction with the new classification. The histological groups of urothelial carcinoma evolved. Grading remained the same, despite controversy among European urologists. Substaging of pT1 tumours is recommended for the first time, and the ICCR has made recommendations on how to report this. Furthermore, worldwide advice has been published on the use of immunohistochemistry, and recommendations have been made to try to standardize the handling of bladder cancer from a histopathological point of view. At a molecular level, bladder cancer groups have been stratified, and an upcoming molecular classification permits a novel view of this malignancy. This review will try to summarize the most important changes.

The impact of subdividing the "atypical" category for urinary cytology on patient management.

INTRODUCTION: The purpose of the study is to determine the impact of subdividing the "atypical" cytology interpretation into two groups: Atypical urothelial cells of uncertain significance (AUC-US) and Atypical urothelial cells suspicious for high-grade urothelial carcinoma (AUC-H/SHGUC), on management of patients with no prior history of UC. MATERIALS AND METHOD: This is a retrospective study of "atypical" urine cytology with subsequent tissue examination occurring within six months. Cytology reports with "atypical" interpretation were reclassified into AUS-UC and AUC-H based on morphologic features identified by the Johns Hopkins system and the Paris system for urine cytology. Follow-up and categorical outcomes were compared between the reclassified AUC-US and AUC-H groups. RESULTS: There was no significant difference (P < 0.4539) in the rate of cytology follow-up, the follow-up cytology result (P < 0.1845), or time between follow-up cytologies (P < 0.0869) between the reclassified atypical group of AUC-H and AUC-US. There was a significant association (P < 0.0001) of rate of malignancy with the reclassified AUC-H (87.18%) compared to the AUC-US (58.68%) groups. CONCLUSION: There was no difference in follow-up between the AUC-H and AUC-US, however there was a difference in the rates of malignancy in the two groups. The AUC-H group is similar to the SHGUC group of the Paris system and can be considered as such, whereas the AUC-US group should continue to be considered atypical. We conclude that reclassification of the "atypical" category into AUC-US and AUC-H/SHGUC can reduce the rate of atypia and help in focused follow-up and targeted management. Diagn. Cytopathol. 2016;44:477-482. © 2016 Wiley Periodicals, Inc.

Classification of samples into two or more ordered populations with application to a cancer trial.

In many applications, especially in cancer treatment and diagnosis, investigators are interested in classifying patients into various diagnosis groups on the basis of molecular data such as gene expression or proteomic data. Often, some of the diagnosis groups are known to be related to higher or lower values of some of the predictors. The standard methods of classifying patients into various groups do not take into account the underlying order. This could potentially result in high misclassification rates, especially when the number of groups is larger than two. In this article, we develop classification procedures that exploit the underlying order among the mean values of the predictor variables and the diagnostic groups by using ideas from order-restricted inference. We generalize the existing methodology on discrimination under restrictions and provide empirical evidence to demonstrate that the proposed methodology improves over the existing unrestricted methodology. The proposed methodology is applied to a bladder cancer data set where the researchers are interested in classifying patients into various groups.

Neural network-based assessment of prognostic markers and outcome prediction in bilharziasis-associated bladder cancer.

In this paper the potential value of two prognostic factors, namely, bilharziasis status and tumor histological type, is investigated in relation to their abilities to predict disease progression and outcome of patients with bladder cancer, using radial basis function (RBF) neural networks. The bladder cancer data set is described by eight clinical and pathological markers. Two outcomes are of interest: either a patient is alive and free of disease or the patient is dead within five years of diagnosis. Three hundred and twenty-one (321) patients are involved in this retrospective study, 83.5% of whom had been confirmed with bilharziasis history. Selected marker subsets are examined to improve the outcome predictive accuracy and to evaluate the effects of the assessed prognostic factors on such outcome. The highest predictive accuracy for patients with bladder adenocarcinoma, as obtained from the RBF network, is found to be 85% with one subset of markers. The predictive analysis shows that bilharziasis history and patients' histology type are both important prognostic factors in prediction and, for each histology type, different marker combinations with significant characteristics have been observed.

Assessment of basal cell status and proliferative patterns in flat and papillary urothelial lesions: a contribution to the new WHO classification of the urothelial tumors of the urinary bladder.

In 1999, the World Health Organization (WHO) published a new classification of papillary urothelial tumors of the urinary bladder. Intended to represent a reproducible, easy-to-use classification system that better separates patients with true malignancies (bladder cancer) from those patients who are at an increased risk for developing bladder cancer, problems in the differential diagnosis of various lesions remained. Probably the most critical distinction is between papillomas, papillary urothelial neoplasms of low malignant potential (lmp), and grade I papillary carcinomas. Conversely, problems in the distinction between reactive atypia, atypia of unknown significance, and dysplasia, as well as the distinction of dysplasia from carcinoma in situ (CIS), are unresolved. Whether urothelial basal cell status assessment on hematoxylin and eosin-stained slides completed by cytokeratin immunohistochemistry with anticytokeratin clone 34betaE12 may help to improve some of the previously mentioned diagnostic dilemmas was investigated. Basal cell status assessment was helpful in the differentiation between dysplasia and CIS. In dysplasia, CK IHC showed a predominantly basal labeling pattern, whereas in CIS, labeling of all urothelial layers was seen. Basal cell status assessment could separate 2 groups of pTa GIb papillary carcinoma. Group 1 with a continuous basal CK labeling and a low MIB-1 labeling index (LI) was compared with group 2, with a diffuse labeling pattern and a significantly higher MIB-1 LI. Whether group 1 carcinomas should better be assigned to the group of papillary urothelial neoplasms of lmp is discussed.

The use of the decision tree technique and image cytometry to characterize aggressiveness in World Health Organization (WHO) grade II superficial transitional cell carcinomas of the bladder.

The aggressiveness of human bladder tumours can be assessed by means of various classification systems, including the one proposed by the World Health Organization (WHO). According to the WHO classification, three levels of malignancy are identified as grades I (low), II (intermediate), and III (high). This classification system operates satisfactorily for two of the three grades in forecasting clinical progression, most grade I tumours being associated with good prognoses and most grade III with bad. In contrast, the grade II group is very heterogeneous in terms of their clinical behaviour. The present study used two computer-assisted methods to investigate whether it is possible to sub-classify grade II tumours: computer-assisted microscope analysis (image cytometry) of Feulgen-stained nuclei and the Decision Tree Technique. This latter technique belongs to the Supervised Learning Algorithm and enables an objective assessment to be made of the diagnostic value associated with a given parameter. The combined use of these two methods in a series of 292 superficial transitional cell carcinomas shows that it is possible to identify one subgroup of grade II tumours which behave clinically like grade I tumours and a second subgroup which behaves clinically like grade III tumours. Of the nine ploidy-related parameters computed by means of image cytometry [the DNA index (DI), DNA histogram type (DHT), and the percentages of diploid, hyperdiploid, triploid, hypertriploid, tetraploid, hypertetraploid, and polyploid cell nuclei], it was the percentage of hyperdiploid and hypertetraploid cell nuclei which enabled identification, rather than conventional parameters such as the DI or the DHT.

The effect of inaccuracies in death certification and coding practices in the European Economic Community (EEC) on international cancer mortality statistics.

An investigation into the effect of national death certification and coding practices on published mortality statistics in eight EEC countries is reported. Doctors in each country were asked to complete specimen death certificates for a bank of written case histories. Certificates from each country were coded by their own offices and then by a WHO reference centre. Within and between countries, discrepancies occurred both in the doctors' diagnoses and in the codes assigned to certificates. At an international level these differences had serious implications for the comparability of mortality data for cancers of the cervix and uterus, and for mesotheliomas.