Prostate Cancer Imaging

Division Of Basic Sciences - Nci

Investigators

Linked publications & trials

Abstract

Prostate cancer is the 2nd leading cause of cancer death in males. As a result of screening with serum prostate specific antigen (PSA) there has been a dramatic increase in the number of men diagnosed with prostate cancer. The diagnosis is being made at a younger age yet the morbidity of the standard treatment such as surgery and radiation---remain unchanged. Thus, men with the diagnosis of prostate cancer are often overtreated for their disease and will live with the consequences of this treatment for many years, greatly affecting the Quality of Life Years (QALY). The ultimate answer to this dilemma is serum biomarkers that identify lethal cancers but not incidental cancers but this is unlikely to occur soon. In the meantime, methods of localizing prostate cancer and treating them with minimally invasive therapy would dramatically lessen the morbidity associated with widespread screening and the overdiagnosis/overtreatment of prostate cancer. A variety of imaging methods have been developed and we are exploring their role in localizing early prostate cancer. The MIP has partnered with the Urologic Oncology Branch to develop new imaging methods to be coupled with minimally invasive treatment methods which include RF/laser ablation, targeted radiation and photoimmunotherapy. The MIP is engaged in a number of pre-clinical studies in prostate cancer. We have been investigating a variety of targeted imaging agents. Initially, we have evaluated several antibodies and antibody fragments against PSMA (3TC, J591) in animal models of prostate cancer. Recently we began using PSMA PET based on small molecules that target PSMA. We are seeking to understand at a molecular level what controls the expression of PSMA in a cancer cell. Metabolic studies of prostate cancer may have implications for therapy. We are investigating PET in animal models of prostate cancer to detect differences in metabolism between the androgen-sensitive and insensitive state. Moreover, we are launching a program in C13 hyperpolarization of pyruvate which will enable non invasive monitoring of metabolism Clinical Trials The MIP has been studying prostate cancer imaging in humans since its inception. We have developed analytic tools in conjunction with a CRADA with Philips Medical Systems. We have demonstrated that a multiparametric approach improves the specificity of 3T MRI for prostate cancer. However, there remain significant limitations in the sensitivity and specificity of 3T MRI. After patients undergo prostatectomy and their specimens are available for review it is clear that less than 40% would be amenable for focal therapy based on being single, well circumscribed lesions that are visible on MRI. We have designed a customized prostate mold which is in use in all patients undergoing surgery at NCI. It enables the imaging to be directly correlated with the pathologic specimen. Therefore, working with Philips Medical Systems we have designed an US-MR fusion system that takes the data from the 3T MRI and fuses it to the real time ultrasound image. This product was released for sale in 2013 under the name UroNav. Biopsy and interventional procedures can then be performed under MR guidance using the ultrasound. This device has been used successfully in over 2000 patients at NCI. A dog study showed the accuracy was about 3-4mm but was much better than cognitive biopsies. We are currently using a similar platform to direct focal laser ablation (FLA) of suitable prostate lesions (i.e. ones that are well demarcated, low Gleason score and could otherwise be watched. We are also testing HIFU in animal models. Computer aided diagnosis and machine learning are new areas of research for the MIP. Significant progress has been made in designing new AI algorithms to help radiologists identify cancers and characterize them. We have extended these AI algorithms to encompass pathologic analysis of biopsy specimens in the hopes of better predicting high risk patients. We are developing novel PET agents that might be more specific for prostate cancer, especially in detecting metastases. For instance, we completed a PET study using 11C-Acetate and 18F-ACBC, an agent associated with amino acid transport which has shown success in localizing recurrent prostate cancer. Ultimately we wish to combine PET-MR studies and conduct minimally invasive therapies after US fusion. We completed a trial of the agent, F18-DCFBC which is a first generation low molecular weight tracer directed at PSMA. We are performing PET studies with a PSMA targeted PET agent. We are conducting trials using F18 DCFPyl, a second generation agent directed at PSMA. This agent has already shown great potential to characterize the disease status even when post treatment PSA levels are quite low. We are hoping to use cold PSMA injected retrograde in the salivary ducts to block uptake in the salivary glands. This will be useful in preserving salivary function when PSMA-directed therapies such as lutetium 177 PSMA, are used to treat metastatic castrate resistant prostate cancer. We are also generating computer aided diagnosis (CAD) devices and algorithms that could automate the diagnosis of prostate cancer. As part of a FLEX grant we have tested numerous other PET agents including F-DHT, arachidonic acid, CXCR4, FDG and F Glutamine and F Bombesin in order to select better agents for imaging prostate cancer. Moreover, since PSMA is so important we have begun basic research to evaluate the expression of PSMA and what controls it. For instance, we found PSMA is one a small group of genes that is regulated by numerous co-factors. We have improved the understanding of FOLH1 regulation which is responsible for PSMA expression. Also we have demonstrated alterations in metabolism related to specific genetic mutations (RB1, TP53) in adenocarcinomas. ADT slow release implants are also introduced into tumors to reduce need for systemic ADT especially during radiation therapy. Artificial intelligence methods are being used to improve diagnostics with imaging and digital pathology. A major project is underway with the Joint Pathology Center at Walter Reed Military Medical Center to evaluate 30 years of digital pathologic specimens of prostatectomy specimens using AI.