Future Trends in Clinical-trial Imaging
This is the final article in a series of three providing a primer for radiologists and imaging professionals interested in clinical trials. Over the past two decades, there has been a significant evolution in the methodology of clinical trials, specifically as it applies to the use of medical imaging as a surrogate endpoint or biomarker. While the use of medical imaging has now garnered mainstream acceptance in clinical-trial arenas to make trials more effective and accurate, the actual use of imaging in trials remains in its infancy. In order to satisfy regulatory requirements and access the most widely available technology, imaging contract-research organizations (CROs) have primarily advocated the use of conventional modalities and methods, as they are easily translatable from the clinical sphere to research protocols. As mainstream acceptance has widened, however, investigators have begun to look into new modalities, newer protocols, and new methods for using medical imaging in clinical trials. Hardware As most current clinical trials involve oncology, the workhorse modalities are CT and MRI (due to the validation and acceptance of response-evaluation criteria in solid tumors). CT technology has evolved to allow isometric voxel imaging and reproducible 3D imaging, and MRI allows evaluation of a broad range of imaging elements, from metabolism and physiology to tissue microstructure. CT: The main developments in CT have been related to the introduction of multidetector spiral CT. Multidetector imaging allows multiplanar and 3D rendering of anatomical structures with a level of detail that is on par with that of other planar modalities—conventional angiography, for example, when compared with CT angiography. The evolution of CT technology has produced improvements in multiple areas, including better spatial resolution, reduced radiation dose, in-vivo high-resolution CT imaging, flat-panel–detector CT, and dual-energy CT. As these technologies are vetted and gain acceptance in clinical practice, they will become involved in protocols for trial imaging. MRI: The evolution of MRI technology has seen the use of both 1.5T and 3T magnets become conventional. For body applications, while 1.5T remains the current standard, the benefits of increased resolution at 3T ultimately will translate to trial imaging. The exquisite imaging obtained on 3T units for neuroimaging and the reproducibility available for cardiac imaging have already caused 3T MRI to make its way into limited clinical trials. Ultrahigh-field (7T) imaging systems are currently in research use, typically for neuroimaging applications, but these are unlikely to find acceptance as a clinical technology (and, consequently, in clinical trials) due to their high overhead and limited scope of utility for clinical applications. Since the introduction of multichannel coils, eight and 16-channel devices have become increasingly common, and modern systems are designed to accommodate upgrades to 24 or 32 channels (or more). Devices that are specific to laboratory research have been taken up to 100 channels. The higher-channel devices allow research protocols to focus on specific organs or specific disease processes, permitting imaging charters to be more precise in the impact of drugs being studied. Integrated modalities: These (especially PET/CT and PET/MRI) are a particular focus for oncology applications. While PET, as a stand-alone modality, has been validated for use in clinical trials, radiologists have limited its use primarily to diagnosis (due to the lack of anatomical data). Integrating PET and CT provides the anatomical detail that helps imaging CROs to decrease the level of adjudication required across time points. Manufacturers have surmounted incredible challenges in coupling the high magnetic fields of the MRI magnets with PET equipment. As this challenge is overcome, PET/MRI will make the natural progression from the clinical sphere to trial methodology, promoting a more accurate understanding of the relationship of metabolic and structural data through a higher spatial resolution than is possible with PET/CT. Software Improvements in software relate primarily to management of trials and trial imaging. This field also has witnessed an explosion in the use of software, on dedicated workstations, that allows improved visualization of imaging data. Trial-management software: Clinical trials were once managed using stand-alone, proprietary information systems. This reliance resulted in the use of legacy systems that are slow, outdated, and ineffective. Newer, improved technology can reduce trial turnaround times tenfold. Image collection: The key areas of differentiation for newer software are improved processes for image collection and image submission, as well as for removal of individuals’ identification. With the use of distributed networks, multiple sites can submit real-time data to a central imaging repository—with unparalleled security. The key ingredients of newer imaging CRO systems include tools that allow monitoring of sites, acquisitions, and quality-assurance (QA) and quality-control (QC) processes in real time. These systems allow sponsors, project managers, and other key personnel full access to images and information. Web-based viewers: For radiologists interpreting studies for clinical trials, these newer systems allow real-time monitoring of studies, as well as improving the activity of interpretation. Given the instantaneous nature of these systems, radiologists can participate in a complex protocol, reviewing images on patients as they are acquired and providing QC. For example, if a sequence in a complex protocol is performed incorrectly, a radiologist could be involved while the patient is on the table (in a different location), making adjustments accordingly. In addition, the use of decentralized Web-based viewers allows geographically unrestricted interpretation. While meeting FDA standards and maintaining good clinical practice, these newer systems allow radiologists with subspecialty expertise in geographically disparate locations to participate in trials. Involving subspecialty radiologists improves the overall quality of interpretation. This ultimately leads to fewer patients being dropped from trial protocols due to incorrect understanding of imaging findings, and it improves the characterization of response (versus nonresponse) to the use of drugs or devices. High-quality interpretations by subspecialty radiologists increase the efficiency—and decrease the cost—of trials. Data integration and reporting: As the volume of data generated by advanced scanners proliferates, workstations and clinical-trial software need to be able to integrate, analyze, and report the data better. Next-generation management software provides new methods for data integration across the clinical-trial platform, allowing better integration of the medical images as they relate to patients enrolled in the trial. In addition, it offers improved tools for analysis and reporting that provide a broader scope, while maintaining the ability to find an individual treatment at a specific time. Study-specific workflow configuration: As trial protocols evolve around newer technologies and better methodologies, clinical-trial software needs to allow for sponsor-driven, imaging CRO controlled specific workflow configurations. As legacy modalities from clinical practice and new modalities in the trial practice gain wider use, many protocols need to be configured to address particular issues regarding certain tools. For example, newer PET radiopharmaceuticals have half-lives measured in minutes. As this mandates that a clinical-trial acquisition site be near a cyclotron, it is of great benefit to be able to configure workflow prior to transmitting test images or enrolling a patient. Storage: Conventionally, imaging CROs have used paper-based storage or, more recently, limited hard-drive storage (with redundant backup) at a single physical location. Next-generation imaging CROs have embraced the use of off-site redundant SAN technology. This allows for the utmost security of data. Other advances: As popular services such as Twitter have gained market share, systems that incorporate similar real-time notifications in their software have become available to alert project managers or QA/QC personnel of potential problems. This allows imaging CROs to capture cleaner data and to decrease the number of patients/cases that need to be discarded due to improper acquisition or screening. Workstations: While using subspecialty radiologists is a key to getting better data in clinical trials, these radiologists, in their clinical work, have begun to use specialized workstations to interrogate the data generated by improved modalities. These workstations allow radiologists to provide more detailed analyses of structures affected by drugs or devices. Other workstation technologies that are raising the interest of trial-methodology architects include computer-aided detection, in-vivo modeling, fusion software, and physiologic measurement. The integration of medical imaging into clinical trials is now validated and accepted. Legacy imaging CROs have helped to build this acceptance, but next-generation providers have begun to employ tools that promise to improve, accelerate, and reduce costs for the trials that bring drugs and devices to the marketplace. As shown by the development of the Metrics Champion Consortium (www.metricschampion.org), imaging CROs are trying to standardize clinical trials. As we integrate newer modalities, technologies, and processes into clinical trials, we will improve our ability to bring drugs and devices to the public in record time. Amit Mehta, MD, FRCP, is a vascular and interventional radiologist with the South Texas Radiology Group in San Antonio and is vice president of an imaging contract-research organization; dramitmehta@intrinsiccro.com.