Validation of the COBRA nomogram for the prediction of cancer specific survival in patients treated with radical cystectomy for bladder cancer: An international wide cohort study

Using machine learning for predicting cancer-specific mortality in bladder cancer patients undergoing radical cystectomy: a SEER-based study

BACKGROUND

Accurately assessing the prognosis of bladder cancer patients after radical cystectomy has important clinical and research implications. Current models, based on traditional statistical approaches and complex variables, have limited performance. We aimed to develop a machine learning (ML)-based prognostic model to predict 5-year cancer-specific mortality (CSM) in bladder cancer patients undergoing radical cystectomy, and compare its performance with current validated models.

METHODS

Patients were selected from the Surveillance, Epidemiology, and End Results database and the First Affiliated Hospital of Sun Yat-sen University for model construction and validation. We used univariate and multivariate Cox regression to select variables with independent prognostic significance for inclusion in the model's construction. Six ML algorithms and Cox proportional hazards regression were used to construct prediction models. Concordance index (C-index) and Brier scores were used to compare the discrimination and calibration of these models. The Shapley additive explanation method was used to explain the best-performing model. Finally, we compared this model with three existing prognostic models in urothelial carcinoma patients using C-index, area under the receiver operating characteristic curve (AUC), Brier scores, calibration curves, and decision curve analysis (DCA).

RESULTS

This study included 8,380 patients, with 6,656 in the training set, 1,664 in the internal validation set, and 60 in the external validation set. Eight features were ultimately identified to build models. The Light Gradient Boosting Machine (LightGBM) model showed the best performance in predicting 5-year CSM in bladder cancer patients undergoing radical cystectomy (internal validation: C-index = 0.723, Brier score = 0.191; external validation: C-index = 0.791, Brier score = 0.134). The lymph node density and tumor stage have the most significant impact on the prediction. In comparison with current validated models, our model also demonstrated the best discrimination and calibration (internal validation: C-index = 0.718, AUC = 0.779, Brier score = 0.191; external validation: C-index = 0.789, AUC = 0.884, Brier score = 0.137). Finally, calibration curves and DCA exhibited better predictive performance as well.

CONCLUSIONS

We successfully developed an explainable ML model for predicting 5-year CSM after radical cystectomy in bladder cancer patients, and it demonstrated better performance compared to existing models.

A Semi-Supervised Multi-Region Segmentation Framework of Bladder Wall and Tumor with Wall-Enhanced Self-Supervised Pre-Training

Bladder cancer is a prevalent and highly recurrent malignancy within the urinary tract. The accurate segmentation of the bladder wall and tumor in magnetic resonance imaging (MRI) is a crucial step in distinguishing between non-muscle-invasive and muscle-invasive types of bladder cancer, which plays a pivotal role in guiding clinical treatment decisions and influencing postoperative quality of life. The performance of data-driven methods is highly dependent on the quality of the annotations and datasets, however the amount of high-quality annotated data is very limited given the difficulty of professional radiologists to distinguish the mixed regions between the bladder wall and the tumor. The performance of the data-driven approach is highly dependent on the quality of the annotation and datasets, Therefore, in order to alleviate these problems and take full advantage of the potential of limited annotated and unlabeled data, we designed a semi-supervised multi-region framework for bladder wall and tumor segmentation. Our framework incorporates wall-enhanced self-supervised pre-training, designed to enhance discrimination of the bladder wall, and a semi-supervised segmentation network that utilizes both limited high-quality annotated data and unlabeled data. Contrast consistency and reconstruction observation losses are introduced to constrain the model to enhance the bladder walls, and adaptive learning rate and post-processing techniques are implemented to further improve segmentation performance. Extensive experimental validation demonstrated that our proposed method achieves promising results in the segmentation of both the bladder wall and the tumor. The average Dice Similarity Coefficients (DSCs) of the proposed method for the bladder wall and tumor were 0.8351 and 0.9175, respectively. Visualization results indicated that our method can effectively reduce excessive segmentation artifacts outside the bladder, and improve the clinical significance of the segmentation results.

Identification of potential DNA methylation biomarkers related to diagnosis in patients with bladder cancer through integrated bioinformatic analysis

BACKGROUND

Bladder cancer (BLCA) is one of the most common malignancies among tumors worldwide. There are no validated biomarkers to facilitate such treatment diagnosis. DNA methylation modification plays important roles in epigenetics. Identifying methylated differentially expressed genes is a common method for the discovery of biomarkers.

METHODS

Bladder cancer data were obtained from Gene Expression Omnibus (GEO), including the gene expression microarrays GSE37817( 18 patients and 3 normal ), GSE52519 (9 patients and 3 normal) and the gene methylation microarray GSE37816 (18 patients and 3 normal). Aberrantly expressed genes were obtained by GEO2R. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were analyzed using the DAVID database and KOBAS. Protein-protein interactions (PPIs) and hub gene networks were constructed by STRING and Cytoscape software. The validation of the results which was confirmed through four online platforms, including Gene Expression Profiling Interactive Analysis (GEPIA), Gene Set Cancer Analysis (GSCA), cBioProtal and MEXPRESS.

RESULTS

In total, 253 and 298 upregulated genes and 674 and 454 downregulated genes were identified for GSE37817 and GSE52519, respectively. For the GSE37816 dataset, hypermethylated and hypomethylated genes involving 778 and 3420 genes, respectively, were observed. Seventeen hypermethylated and low expression genes were enriched in biological processes associated with different organ development and morphogenesis. For molecular function, these genes showed enrichment in extracellular matrix structural constituents. Pathway enrichment showed drug metabolic enzymes and several amino acids metabolism, PI3K-Akt, Hedgehog signaling pathway. The top 3 hub genes screened by Cytoscape software were EFEMP1, SPARCL1 and ABCA8. The research results were verified using the GEPIA, GSCA, cBioProtal and EXPRESS databases, and the hub hypermethylated low expression genes were validated.

CONCLUSION

This study screened possible aberrantly methylated expression hub genes in BLCA by integrated bioinformatics analysis. The results may provide possible methylation-based biomarkers for the precise diagnosis and treatment of BLCA in the future.

Development of a Pocket Nomogram to Predict Cancer and Disease Specific Survival After Radical Cystectomy For Bladder Cancer: The CRAB Nomogram

OBJECTIVES

To develop an easy tool to predict cancer specific (CSS) and disease-free survival (DFS) in patients with bladder cancer treated with radical cystectomy.

METHODS

Data from a consecutive series of 2395 patients with primitive or progression to muscle invasive bladder cancer (MIBC) undergone to radical cystectomy and lymph nodes dissection in 5 centers were evaluated. Using Cox proportional hazards analyses, the Cancer of the bladder risk assessment (CRAB) nomogram was generated. Accuracy of the nomogram was evaluated by Harrell's C test. Internal validation of the model was performed using 200 bootstraps.

RESULTS

Median age was 66 (IQR 58/73) years; 612/2395 (26%) patients presented an advanced pathological stage (≥pT3a); 478/2395 (20%) presented positive lymph nodes. Overall, 729/2395 (30%) presented local or distant recurrence with a median DFS of 42 (IQR 14/89) months. Overall, 642/2395 (27%) died of bladder cancer with a median follow up of 48 (IQR 22/92) months. On univariate Cox proportional hazards analyses, age, stage, and lymph nodes density were a significant predictor of 3 and5 years CSS and DFS. Accuracy of the CRAB nomogram was 0.73 and 0.71 respectively.

CONCLUSION

CRAB nomogram can be a practical and easily applicable tool that may help urologists to classify the long-term CSS and DFS of patients treated with radical cystectomy and to predict the oncological outcome.

Surface-Enhanced Raman Spectroscopy of Pretreated Plasma Samples Predicts Disease Recurrence in Muscle-Invasive Bladder Cancer Patients Undergoing Neoadjuvant Chemotherapy and Radical Cystectomy

Objective

To explore the value of surface-enhanced Raman spectroscopy analysis of pretreated plasma samples in prediction of bladder cancer (BCa) recurrence after neoadjuvant chemotherapy (NAC) and radical cystectomy (RC).

Patients and Methods

SERS was used to analyze plasma samples collected before biopsy and treatment in BCa patients undergoing NAC and RC. The value of clinicopathological parameters and distinctive SERS peaks in the prediction of disease recurrence were analyzed in Cox regression proportional hazard analysis and Log rank test. Principal component analysis and linear discriminant analysis (PCA-LDA) were employed to process spectral data and construct diagnostic algorithms.

Results

A total of 88 patients with 440 plasma SERS spectra were collected. The SRES spectra from recurrent patients were compared with patients who remained recurrence free. The SERS demonstrated higher levels of circulating free nucleic acid components in recurrent population, which is represented by significantly higher intensities at SERS peaks of 725 cm⁻¹, 1328 cm⁻¹ and 1455 cm⁻¹. The SERS also detected significantly lower levels of tryptophan shown as lower significantly intensities at the 1558 cm⁻¹, which is proved to be an independent predictor of BCa recurrence. The addition of SERS peaks of 1558 cm⁻¹ to classic clinicopathological predictors including pathological tumor stage, lymph node metastasis and pathological downstaging can significantly enhance the power of the predictive model from 0.66 to 0.76 in the area under curve (AUC) of receiver operating characteristic (ROC) curves. Meanwhile, the PCA-LDA diagnostic model based on SERS spectra reveals a high accuracy of 85.2% in prediction of disease recurrence and the AUC of 0.92 in the ROC curve. When validated in the leave-one-out cross-validation method, the accuracy of the model remained 84.1%.

Conclusion

We show that SERS analysis of plasma before NAC treatment can accurately classify patients with different risks of disease recurrence after surgery and improve the power of clinicopathological predictive models, thus refining clinical decision-making.