Genetically engineered neural networks for predicting prostate cancer progression after radical prostatectomy

Indicators of pathologic stage of prostate cancer and their use in clinical practice

Pathologic stage is the most reliable means of predicting the likelihood of curable prostate cancer at the time of definitive treatment. Its prediction is of the greatest importance to individuals with clinically localized disease, principally because of the therapeutic and prognostic implications. Multivariate models integrating variables that can be derived from clinical and pathologic assessment have been shown to be reliable and useful in urologic practice. Among these variables, the combination of clinical stage, serum PSA, and biopsy Gleason score provides reliable assessment of the risk for extraprostatic disease that can be used readily for counseling individual patients. Other biopsy-derived parameters may contribute additional information, but their value in multivariate analysis has not been validated in a multi-institutional setting. The development of new prognostic markers is a priority objective in current research to distinguish patients in whom cancer cannot be controlled by surgical treatment. For patients undergoing radical prostatectomy, definitive pathologic stage certainly will remain an important prognostic factor; therefore, clinical practice will continue to be determined by its accurate prediction.

Pathologic assessment of the surgical specimen

Despite the wealth of information obtained by conventional histology, long-term studies are needed to provide novel information on the correlation of pathologic findings with prognosis. Findings need to be correlated not only with PSA progression but with the more clinically important parameters of distant metastases and survival. Although conventional histology still will have a role in the evaluation of prostate cancer at radical prostatectomy and its correlation with outcome, it undoubtedly will be augmented by newer techniques. These developments must be approached critically and rationally to determine whether they provide additional prognostic information beyond that currently available using more conventional parameters.

Standardization, analytical validation, and quality control of intermediate endpoint biomarkers

Standardized processes should be used in the identification and development of intermediate endpoint biomarkers (IEB) for the prediction of patient-specific disease outcomes. Using our own experiences, we outline some of our standardized processes. Using computer-assisted image analysis, we developed a new biomarker of genetic instability, termed quantitative nuclear grade (QNG). The QNG biomarker is derived using nuclear images analyzed from the tumor areas of Feulgen-stained 5-microm biopsy or radical prostatectomy tissue sections. From the variances of 41 to 60 different nuclear size, shape, and chromatin organization features, a QNG solution is computed using either logistic regression or artificial neural networks. QNG can then be used as an input for models that solve for a patient-specific probability to accurately predict disease outcomes. Preoperatively, QNG predicted both the pathologic stage and progression of prostate cancer using biopsies (P <0.0001). Postoperatively, QNG proved extremely valuable in the prediction of biochemical progression using radical prostatectomy specimens with more than 10 years of follow-up (P <0.0001). We also demonstrate the identification of novel, differentially expressed, prostate cancer genes using RNA fingerprinting methods and the clinical utility of testing for these genes in both blood and tissue samples. Also illustrated is the improvement of serum biomarker performance by combining molecular forms of PSA with new biomarkers. In conclusion, the development of new IEBs requires planning based upon an understanding of the molecular pathogenesis of disease. IEB selection and clinical evaluation should employ standardized methods of testing and validation, followed by publication. QNG is 1 example of a new, highly predictive, IEB for prostate cancer that has been developed using these processes.

Diagnosis of localized prostate cancer: 10 years of progress

Advances in the diagnosis of early stage disease, and particularly the introduction of prostate-specific antigen (PSA) testing, have had a dramatic effect on the presentation and clinical management of prostate cancer during the past ten years. As a result, there have been significant epidemiological changes in countries where early diagnosis is recommended. The importance of PSA testing for the diagnosis of localized prostate cancer has become well established in clinical practice and this is reflected by improved outcomes from definitive treatment. The contribution of PSA-related parameters and molecular forms of PSA both to cancer detection and prediction of pathological stage continue to be explored. Concerns about the reliability of the standard sextant biopsy technique for cancer detection relate to the need for re-biopsy in a growing number of patients with negative biopsies and an increasing proportion of patients with low volume, multifocal disease. In men with cancer, additional prognostic information can be derived from biopsy findings, with important therapeutic implications. This relates also to the need for reliable markers indicating pathological stage and risk of progression. The opportunities for the prevention of prostate cancer have grown with improved understanding of its biology and the genetic basis of the early steps associated with malignant transformation. In the future, the need for therapeutic intervention is likely to be most influenced by successful prevention strategies.

Intraoperative cavernous nerve stimulation during nerve sparing radical prostatectomy: how and when?

Although nerve-sparing prostatectomy is widely practiced, the results with respect to potency preservation often do not meet expectations. The concept of intraoperative cavernous nerve stimulation is rational. Recent data that link the response to sildenafil after prostatectomy to bilateral nerve sparing has increased the importance of optimizing nerve sparing. The cavernous nerves are often difficult to visualize and may have a variable course. A tumescent response to nerve stimulation can be consistently demonstrated. The response may be subtle, and characterized by a minimal increase in penile circumference and blood flow. Immediately after prostectomy, proximal nerve stimulation identifies whether neural continuity has been maintained, and is predictive of recovery of erectile function. A new device, the Cavermap, has been developed to permit intraoperative nerve stimulation with tumescence monitoring. An initial phase 2 and subsequent phase 3 single blinded, randomized, multicenter study that compared Cavermap-assisted prostatectomy with conventional nerve sparing demonstrated a significant benefit in terms of the duration of nocturnal tumescence by Rigiscan (Timm Medical Technologies, Eden Prairie, Minnesota) at 1 year. Other approaches are being explored, including sural nerve grafting, use of nerve stimulation during cystectomy or abdominal-perineal resection, and direct corpus cavernosum pressure monitoring during nerve stimulation. These approaches warrant further evaluation.