1
|
A novel self-learning framework for bladder cancer grading using histopathological images. Comput Biol Med 2021; 138:104932. [PMID: 34673472 DOI: 10.1016/j.compbiomed.2021.104932] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 06/15/2021] [Revised: 10/07/2021] [Accepted: 10/07/2021] [Indexed: 11/23/2022]
Abstract
In recent times, bladder cancer has increased significantly in terms of incidence and mortality. Currently, two subtypes are known based on tumour growth: non-muscle invasive (NMIBC) and muscle-invasive bladder cancer (MIBC). In this work, we focus on the MIBC subtype because it has the worst prognosis and can spread to adjacent organs. We present a self-learning framework to grade bladder cancer from histological images stained by immunohistochemical techniques. Specifically, we propose a novel Deep Convolutional Embedded Attention Clustering (DCEAC) which allows for the classification of histological patches into different levels of disease severity, according to established patterns in the literature. The proposed DCEAC model follows a fully unsupervised two-step learning methodology to discern between non-tumour, mild and infiltrative patterns from high-resolution 512 × 512 pixel samples. Our system outperforms previous clustering-based methods by including a convolutional attention module, which enables the refinement of the features of the latent space prior to the classification stage. The proposed network surpasses state-of-the-art approaches by 2-3% across different metrics, reaching a final average accuracy of 0.9034 in a multi-class scenario. Furthermore, the reported class activation maps evidence that our model is able to learn by itself the same patterns that clinicians consider relevant, without requiring previous annotation steps. This represents a breakthrough in MIBC grading that bridges the gap with respect to training the model on labelled data.
Collapse
|
2
|
Liang Q, Zhang G, Li W, Wang J, Sheng S. Comparison of the diagnostic performance of fluorescence in situ hybridization (FISH), nuclear matrix protein 22 (NMP22), and their combination model in bladder carcinoma detection: a systematic review and meta-analysis. Onco Targets Ther 2018; 12:349-358. [PMID: 30643432 PMCID: PMC6317485 DOI: 10.2147/ott.s186065] [Citation(s) in RCA: 8] [Impact Index Per Article: 1.3] [Reference Citation Analysis] [Abstract] [Key Words] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/23/2022] Open
Abstract
Background Emerging studies reported that combination of fluorescence in situ hybridization (FISH) and nuclear matrix protein 22 (NMP22) could increase the sensitivity and specificity of bladder carcinoma (BC) management. Nevertheless, the reports remain inconsistent. This meta-analysis was undertaken to evaluate the diagnostic performance of FISH, NMP22, and their combination model in BC. Materials and methods A systematic literature search was carried out in PubMed, Embase, Cochrane Library, Web of Science, Chinese National Knowledge Infrastructure, and Wanfang database dated up to October 2018. Suitable studies were identified and raw data were extracted. Meta-analysis was conducted to calculate the global sensitivities, specificities, likelihood ratio, diagnostic odds ratio (DOR), and the areas under the summary receiver operating characteristic (SROC) curves for FISH, NMP22, and their combination model, separately. All the meta-analysis estimates were derived using STATA (version 12.0) and MetaDisc (version 1.4) software packages. Results Seven eligible studies were included for analysis. The global sensitivities with 95% CI for FISH, NMP22, and their combination model were 0.79 (95% CI: 0.75–0.83), 0.76 (95% CI: 0.71–0.81), and 0.82 (95% CI: 0.75–0.88); specificities were 0.85 (95% CI: 0.76–0.91), 0.70 (95% CI: 0.55–0.81), and 0.90 (95% CI: 0.70–0.97); DORs were 22.215 (95% CI: 10.695–46.144), 7.365 (95% CI: 3.986–13.610), and 41.940 (95% CI: 13.546–129.853); and the areas under the SROC curves were 0.86 (95% CI: 0.82–0.88), 0.79 (95% CI: 0.76–0.83), and 0.90 (95% CI: 0.87–0.92). Conclusion Our systematic review implied that the diagnostic performance of combination model of FISH plus NMP22 may outperform FISH or NMP22 alone in BC detection.
Collapse
Affiliation(s)
- Qindong Liang
- Department of Clinical Laboratory, Affiliated Hospital & Clinical Medical College of Chengdu University, Chengdu, Sichuan, China,
| | - Guangjie Zhang
- Department of Clinical Laboratory, Chengdu Fifth People's Hospital, Chengdu, Sichuan, China
| | - Wuxian Li
- Department of Clinical Laboratory, Chongqing Health Center for Women and Children, Chongqing, China
| | - Jing Wang
- Department of Blood Transfusion, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China
| | - Shangchun Sheng
- Department of Clinical Laboratory, Affiliated Hospital & Clinical Medical College of Chengdu University, Chengdu, Sichuan, China,
| |
Collapse
|
3
|
March-Villalba J, Panach-Navarrete J, Herrero-Cervera M, Aliño-Pellicer S, Martínez-Jabaloyas J. hTERT mRNA expression in urine as a useful diagnostic tool in bladder cancer. Comparison with cytology and NMP22 BladderCheck Test®. Actas Urol Esp 2018; 42:524-530. [PMID: 29631914 DOI: 10.1016/j.acuro.2018.02.002] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/20/2017] [Revised: 01/31/2018] [Accepted: 02/01/2018] [Indexed: 10/17/2022]
Abstract
INTRODUCTION To study the relationship between quantitative mRNA determination (hTERT) in patients with bladder tumor, history of bladder tumor, and in subjects without a history of this neoplasia. MATERIAL AND METHODS A prospective randomized controlled study with 91 subjects included. The value of mRNA-hTERTN was determined in 63 patients with a history or suspicion of bladder tumor and in 28 controls. Urine samples were sent for evaluation of the mRNA level (hTERT), the cytological study and the NMP22 result. RESULTS Differences were observed in mean hTERTN levels in each of the groups: tumor presence 21.33+/- 40.66, tumor history 2.16+/- 2.67, controls 0.9+/- 1, 75 (p<0.001). In patients with tumor, there was no difference in mean hTERTN levels between the different grades and stages, although there was a tendency: low grade tumor 9.04+/- 16.95, high grade 28.95+/- 48.36 (p=.069), stage Ta 10.33+/- 19.39, T1 17.88+/- 27.14, T2 54.8+/- 74.05 (p=.056). In addition, the sensitivity of hTERTN was superior to that of other test (76%), although specificity and positive and negative predictive values were better for cytology (94%, 88.4% and 72.3% respectively) and NMP22 (88%, 80.6% and 73.3% respectively). CONCLUSIONS hTERTN mRNA levels in urine were higher in patients with bladder tumors compared to patients with a history of bladder tumor and with negative cystoscopy, as well as in the control group. This determination showed a higher diagnostic yield compared with the detection of NMP22 and urinary cytology.
Collapse
|
4
|
Butt E, Ebbing J, Bubendorf L, Ardelt P. Influence of hematuria and infection on diagnostic accuracy of urinary LASP1: a new biomarker for bladder carcinoma. Biomark Med 2017; 11:347-357. [PMID: 28290211 DOI: 10.2217/bmm-2016-0348] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 02/07/2023] Open
Abstract
AIM To further promote the clinical use of urinary LASP1 as biomarker for urothelial carcinoma of the bladder regarding limitations and alternative testing systems. PATIENTS & METHODS Urine stabilization, alternative measurement systems and limitations by erythrocyte contamination and infection were investigated in 246 patients. RESULTS Thimerosal allowed sufficient stabilization. Fluorescence-activated cell sorting analysis was not influenced by presence of erythrocytes or leukocytes and reliably urothelial carcinoma of the bladder but cell counts in specimen were low. Cut-off values of <25 leukocytes and <200 erythrocytes/µl resulted in sensitivity, specificity, positive and negative predictive values of 0.59, 0.80, 0.80 and 0.59, respectively. CONCLUSION Hematuria up to 200 erythrocytes/µl but not presence of leukocytes may be tolerated for this promising marker.
Collapse
Affiliation(s)
- Elke Butt
- Institute for Experimental Biomedicine II, University Clinic of Wuerzburg, Germany
| | - Jan Ebbing
- Department of Urology, University Hospital Basel, Basel, Switzerland
| | - Lukas Bubendorf
- Department of Pathology, University Hospital Basel, Basel, Switzerland
| | - Peter Ardelt
- Department of Urology, University Hospital Basel, Basel, Switzerland
| |
Collapse
|
5
|
Ardelt P, Grünemay N, Strehl A, Jilg C, Miernik A, Kneitz B, Butt E. LASP-1, a novel urinary marker for detection of bladder cancer. Urol Oncol 2012; 31:1591-8. [PMID: 22481019 DOI: 10.1016/j.urolonc.2012.02.002] [Citation(s) in RCA: 19] [Impact Index Per Article: 1.6] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 11/19/2011] [Revised: 02/02/2012] [Accepted: 02/02/2012] [Indexed: 12/28/2022]
Abstract
OBJECTIVES The LIM and SH3 (LASP)-1 protein is a focal adhesion protein that has been linked to oncogenesis. The aim of this study was to evaluate the diagnostic use of the detection of LASP-1 in tumor specimens and in urine for noninvasive detection of transitional cell carcinoma (TCC). MATERIALS AND METHODS Immunohistochemical staining for LASP-1 was performed on 72 archived bladder tumor specimens, and LASP-1 content was measured in 132 spontaneous urine sample sediments by Western blot analysis. RESULTS In the histologic specimens, immunohistochemical staining for LASP-1 showed abundant expression throughout the urothelium of the bladder and ureter. However, modest overexpression of LASP-1 was observed in the TCC specimens. Measurement of the LASP-1 protein concentrations in urinary cell pellets from healthy donors and bladder cancer patients was highly sensitive for the presence of TCC. The cut-off value was determined by receiver operating characteristic (ROC) analysis. When a cut-off value of 1 ng LASP-1/500 μl of urine was used, the sensitivity, specificity, and positive and negative predictive values of the assay were 83.1%, 85.3%, 83.1%, and 80.6%, respectively.The increased urinary LASP-1 content in TCC patients was attributable in part to a specific increase in cell shedding presumably caused by changes in cell adhesion, as confirmed by LASP-1 knockdown. Contamination with erythrocytes above 250 cells/μl and urinary infection gave false positive results and are therefore sample exclusion criteria. CONCLUSIONS In the absence of urinary infection or gross hematuria, urinary LASP-1 level is a promising marker for transitional cell carcinoma.
Collapse
Affiliation(s)
- Peter Ardelt
- Department of Urology, Albert Ludwig University of Freiburg Medical School, Freiburg, Germany.
| | | | | | | | | | | | | |
Collapse
|
6
|
Horstmann M, Patschan O, Hennenlotter J, Senger E, Feil G, Stenzl A. Combinations of urine-based tumour markers in bladder cancer surveillance. ACTA ACUST UNITED AC 2009; 43:461-6. [DOI: 10.3109/00365590903296837] [Citation(s) in RCA: 44] [Impact Index Per Article: 2.9] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/13/2022]
Affiliation(s)
- Marcus Horstmann
- Department of Urology, Eberhard Karls University Tuebingen, Tuebingen, Germany
| | - Oliver Patschan
- Department of Urology, Eberhard Karls University Tuebingen, Tuebingen, Germany
| | - Jörg Hennenlotter
- Department of Urology, Eberhard Karls University Tuebingen, Tuebingen, Germany
| | - Erika Senger
- Department of Urology, Eberhard Karls University Tuebingen, Tuebingen, Germany
| | - Gerhard Feil
- Department of Urology, Eberhard Karls University Tuebingen, Tuebingen, Germany
| | - Arnulf Stenzl
- Department of Urology, Eberhard Karls University Tuebingen, Tuebingen, Germany
| |
Collapse
|
7
|
Yu J, Zhu T, Wang Z, Zhang H, Qian Z, Xu H, Gao B, Wang W, Gu L, Meng J, Wang J, Feng X, Li Y, Yao X, Zhu J. A novel set of DNA methylation markers in urine sediments for sensitive/specific detection of bladder cancer. Clin Cancer Res 2008; 13:7296-304. [PMID: 18094410 DOI: 10.1158/1078-0432.ccr-07-0861] [Citation(s) in RCA: 168] [Impact Index Per Article: 10.5] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 01/12/2023]
Abstract
PURPOSE This study aims to provide a better set of DNA methylation markers in urine sediments for sensitive and specific detection of bladder cancer. EXPERIMENTAL DESIGN Fifty-nine tumor-associated genes were profiled in three bladder cancer cell lines, a small cohort of cancer biopsies and urine sediments by methylation-specific PCR. Twenty-one candidate genes were then profiled in urine sediments from 132 bladder cancer patients (8 cases for stage 0a; 68 cases for stage I; 50 cases for stage II; 4 cases for stages III; and 2 cases for stage IV), 23 age-matched patients with noncancerous urinary lesions, 6 neurologic diseases, and 7 healthy volunteers. RESULTS Despite six incidences of four genes reported in 3 of 23 noncancerous urinary lesion patients analyzed, cancer-specific hypermethylation in urine sediments were reported for 15 genes (P < 0.05). Methylation assessment of an 11-gene set (SALL3, CFTR, ABCC6, HPR1, RASSF1A, MT1A, RUNX3, ITGA4, BCL2, ALX4, MYOD1, DRM, CDH13, BMP3B, CCNA1, RPRM, MINT1, and BRCA1) confirmed the existing diagnosis of 121 among 132 bladder cancer cases (sensitivity, 91.7%) with 87% accuracy. Significantly, more than 75% of stage 0a and 88% of stage I disease were detected, indicating its value in the early diagnosis of bladder cancer. Interestingly, the cluster of reported methylation markers used in the U.S. bladder cancers is distinctly different from that identified in this study, suggesting a possible epigenetic disparity between the American and Chinese cases. CONCLUSIONS Methylation profiling of an 11-gene set in urine sediments provides a sensitive and specific detection of bladder cancer.
Collapse
Affiliation(s)
- Jian Yu
- Cancer Genetics and Gene Therapy Program, The State Key Laboratory for Oncogenes and Related Genes, Shanghai Cancer Institute, Shanghai Jiao Tong University, Shanghai, China
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | |
Collapse
|
8
|
Fatela-Cantillo D, Fernández-Suárez A, Menéndez-López V, Carro Rubias C, Galán JA, García Serrado D. [Influence of the delay in urinary NMP22 preanalytical processing]. Actas Urol Esp 2007; 31:746-51. [PMID: 17902468 DOI: 10.1016/s0210-4806(07)73716-6] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/25/2022]
Abstract
OBJECTIVES To evaluate the influence of the delay in urinary NMP22 preanalytical processing. MATERIAL AND METHODS Twenty-eight voided urine samples were taken: bladder cancer (14), urine tract infections (4), lithiasis (4), healthy volunteers (1), and with other no malignant bladder diseases (5). All samples, were maintained at environment temperature, and were processed according to the stabilization of parts of urine collected at 0, 30, 90 and 150 minutes. Samples were stored at 4 degrees C until its determination. NMP22 was determined with the IMMULITE One analyzer. RESULTS There were no significant differences for NMP22 levels between each different point of time studied. CONCLUSIONS Delay up to 2 hours and a half when we add stabilization solution to urine samples no affects NMP22 results. That thing, might provide more confidence and flexibility on quantitative immunoassays that required urine stabilization.
Collapse
Affiliation(s)
- D Fatela-Cantillo
- Area de Biotecnología, Centro Hospitalario de Alta Resolución Sierra del Segura, EPHAG, La Puerta de Segura-Puente de Génave, Jaén.
| | | | | | | | | | | |
Collapse
|