Collectives of diagnostic biomarkers identify high-risk subpopulations of hematuria patients: exploiting heterogeneity in large-scale biomarker data.

BACKGROUND: Ineffective risk stratification can delay diagnosis of serious disease in patients with hematuria. We applied a systems biology approach to analyze clinical, demographic and biomarker measurements (n = 29) collected from 157 hematuric patients: 80 urothelial cancer (UC) and 77 controls with confounding pathologies. METHODS: On the basis of biomarkers, we conducted agglomerative hierarchical clustering to identify patient and biomarker clusters. We then explored the relationship between the patient clusters and clinical characteristics using Chi-square analyses. We determined classification errors and areas under the receiver operating curve of Random Forest Classifiers (RFC) for patient subpopulations using the biomarker clusters to reduce the dimensionality of the data. RESULTS: Agglomerative clustering identified five patient clusters and seven biomarker clusters. Final diagnoses categories were non-randomly distributed across the five patient clusters. In addition, two of the patient clusters were enriched with patients with 'low cancer-risk' characteristics. The biomarkers which contributed to the diagnostic classifiers for these two patient clusters were similar. In contrast, three of the patient clusters were significantly enriched with patients harboring 'high cancer-risk" characteristics including proteinuria, aggressive pathological stage and grade, and malignant cytology. Patients in these three clusters included controls, that is, patients with other serious disease and patients with cancers other than UC. Biomarkers which contributed to the diagnostic classifiers for the largest 'high cancer- risk' cluster were different than those contributing to the classifiers for the 'low cancer-risk' clusters. Biomarkers which contributed to subpopulations that were split according to smoking status, gender and medication were different. CONCLUSIONS: The systems biology approach applied in this study allowed the hematuric patients to cluster naturally on the basis of the heterogeneity within their biomarker data, into five distinct risk subpopulations. Our findings highlight an approach with the promise to unlock the potential of biomarkers. This will be especially valuable in the field of diagnostic bladder cancer where biomarkers are urgently required. Clinicians could interpret risk classification scores in the context of clinical parameters at the time of triage. This could reduce cystoscopies and enable priority diagnosis of aggressive diseases, leading to improved patient outcomes at reduced costs.

The impact of biomarkers in multivariate algorithms for bladder cancer diagnosis in patients with hematuria.

BACKGROUND: We appraised 23 biomarkers previously associated with urothelial cancer in a case-control study. Our aim was to determine whether single biomarkers and/or multivariate algorithms significantly improved on the predictive power of an algorithm based on demographics for prediction of urothelial cancer in patients presenting with hematuria. METHODS: Twenty-two biomarkers in urine and carcinoembryonic antigen (CEA) in serum were evaluated using enzyme-linked immunosorbent assays (ELISAs) and biochip array technology in 2 patient cohorts: 80 patients with urothelial cancer, and 77 controls with confounding pathologies. We used Forward Wald binary logistic regression analyses to create algorithms based on demographic variables designated prior predicted probability (PPP) and multivariate algorithms, which included PPP as a single variable. Areas under the curve (AUC) were determined after receiver-operator characteristic (ROC) analysis for single biomarkers and algorithms. RESULTS: After univariate analysis, 9 biomarkers were differentially expressed (t test; P < .05). CEA AUC 0.74; bladder tumor antigen (BTA) AUC 0.74; and nuclear matrix protein (NMP22) 0.79. PPP included age and smoking years; AUC 0.76. An algorithm including PPP, NMP22, and epidermal growth factor (EGF) significantly improved AUC to 0.90 when compared with PPP. The algorithm including PPP, BTA, CEA, and thrombomodulin (TM) increased AUC to 0.86. Sensitivities = 91%, 91%; and specificities = 80%, 71%, respectively, for the algorithms. CONCLUSIONS: Addition of biomarkers representing diverse carcinogenic pathways can significantly impact on the ROC statistic based on demographics. Benign prostate hyperplasia was a significant confounding pathology and identification of nonmuscle invasive urothelial cancer remains a challenge.

The use of morphological characteristics and texture analysis in the identification of tissue composition in prostatic neoplasia.

Quantitative examination of prostate histology offers clues in the diagnostic classification of lesions and in the prediction of response to treatment and prognosis. To facilitate the collection of quantitative data, the development of machine vision systems is necessary. This study explored the use of imaging for identifying tissue abnormalities in prostate histology. Medium-power histological scenes were recorded from whole-mount radical prostatectomy sections at x 40 objective magnification and assessed by a pathologist as exhibiting stroma, normal tissue (nonneoplastic epithelial component), or prostatic carcinoma (PCa). A machine vision system was developed that divided the scenes into subregions of 100 x 100 pixels and subjected each to image-processing techniques. Analysis of morphological characteristics allowed the identification of normal tissue. Analysis of image texture demonstrated that Haralick feature 4 was the most suitable for discriminating stroma from PCa. Using these morphological and texture measurements, it was possible to define a classification scheme for each subregion. The machine vision system is designed to integrate these classification rules and generate digital maps of tissue composition from the classification of subregions; 79.3% of subregions were correctly classified. Established classification rates have demonstrated the validity of the methodology on small scenes; a logical extension was to apply the methodology to whole slide images via scanning technology. The machine vision system is capable of classifying these images. The machine vision system developed in this project facilitates the exploration of morphological and texture characteristics in quantifying tissue composition. It also illustrates the potential of quantitative methods to provide highly discriminatory information in the automated identification of prostatic lesions using computer vision.

A computer-based training system for breast fine needle aspiration cytology.

Fine-needle aspiration (FNA) cytology is a rapid and inexpensive technique used extensively in the diagnosis of breast disease. To remove diagnostic subjectivity, a diagnostic decision support system (DDSS) called CytoInform has been developed, based on a Bayesian belief network (BBN) for the diagnosis of breast FNAs. In addition to acting as a DDSS, the system implements a computer-based training (CBT) system, providing a novel approach to breast cytology training. The system guides the trainee cytopathologist through the diagnostic process, allowing the user to grade each diagnostic feature using a set of on-screen reference images as visual clues. The trainee positions a slider on a spectrum relative to these images, reflecting the similarity between the reference image and the microscope image. From this, an evidence vector is generated, allowing the current diagnostic probability to be updated by the BBN. As the trainee assesses each clue, the evidence entered is compared with that of the expert through the use of a defined teaching file. This file records the relative severity of each clue and a tolerance band within which the trainee must position the slider. When all clues in the teaching case have been completed, the system informs the user of inaccuracies and offers the ability to reassess problematic features. In trials with two pathologists of different experience and a series of ten cases, the system provided an effective tool in conveying diagnostic evidence and protocols to trainees. This is evident from the fact that each pathologist only misinterpreted one case and a total of 86%/88% (experienced/inexperienced) of all clues assessed were interpreted correctly. Significantly, in all cases that produced the correct final diagnostic probability, the route taken to that solution was consistent with the expert's solution.