Imaging in Clinical Trials: Endpoints, Biomarkers, and Methods
This article is the second in a series of three providing a background and primer to radiologists and imaging professionals interested in clinical trials. It offers background on regulatory approval and on the use of surrogate endpoints and imaging biomarkers, as well as touching on the administration of medical imaging in clinical trials. Medical imaging has become a significant tool in the development of clinical-trial protocols. Diagnostic imaging enables sponsors to obtain a rapid diagnosis by using truly quantitative assessment, as well as qualitative assessment. Paramount to understanding the utility of medical imaging in clinical trials is an understanding of biomarkers as imaging endpoints. The definitions of these key elements of trial mechanics were developed and distributed by the Biomarkers Definition Working Group of the National Institutes of Health (NIH).1 The key determination point in a clinical trial is called a clinical endpoint. A clinical endpoint specifically refers to a characteristic or variable such as a sign, symptom, laboratory abnormality, or disease that represents one of the targets of the trial. It often represents how an enrolled patient feels, functions, or survives. The results of a clinical trial most often try to represent the number of patients who reach the clinical endpoint during the study, in comparison with the total number of patients who are enrolled. When patients enrolled in the trial reach the clinical endpoint, they are often excluded from other facets of the trial, so the definition and determination of the endpoint are critical. For example, a clinical trial testing the ability of a drug to prevent deep-vein thrombosis could use leg pain as the clinical endpoint. A person enrolled in the trial who develops this symptom would be categorized as having met the clinical endpoint. The use of classic clinical endpoints, however, often can lead to lengthy trials—and, based on design, nonspecific correlation with the true effect of the therapy that is being evaluated. Therefore, clinical-trial methodology experts and researchers began to investigate alternative techniques to assist them in improving the elapsed time, quality, and statistical power of clinical trials This investigation led to the development of surrogate endpoints. Choosing Surrogates To substitute for a clinical endpoint, a surrogate endpoint or surrogate marker is used. In the majority of instances, surrogate endpoints have been shown to decrease the length of trials, as there is more rapid assessment of whether a drug has clinical benefits. An endpoint that is simply correlated with a clinical endpoint that is being studied might not always be a suitable surrogate, as the surrogate is expected to predict the benefit (or lack thereof) that the treatment has on the clinical endpoint.1 The NIH has defined two specific criteria in relation to surrogate endpoints. First, a surrogate endpoint is, specifically, a biomarker intended to substitute for a clinical endpoint.2 ,3 Second, a biomarker is a characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic response to a therapeutic intervention.1 Over the course of the early 1990s, the FDA began approving trials that used a test or result that measured and evaluated a biological process, pathogenic process, or pharmacologic response to a treatment or drug rather than requiring a classic clinical endpoint. An example of a biomarker is prostate-specific antigen (PSA), used in evaluating the response to therapy for prostate cancer. A trial can be designed to use a decrease in PSA (the biomarker) as a surrogate endpoint for response to therapy. With the use of biomarkers and surrogate endpoints over the past decade, researchers have been able to use smaller group sizes to obtain quicker results while maintaining the necessary statistical power.4Imaging Biomarkers Specific to radiology, imaging biomarkers are a subtype of biomarker that enhances the methodology of a clinical trial. An imaging biomarker specifically uses a characteristic that is objectively measured using an imaging technique; this indirectly serves to represent the pharmacological response to the therapy being administered. The benefit of using medical imaging (correctly interpreted) is the ability to reveal subtle changes that demonstrate the effectiveness of therapy—and that are often missed by the more usual subjective methods. In addition, trial data are less tainted by subjectivity, as imaging findings are subject to blinding and are evaluated without contact with the patient.
Figure. Functional MRI demonstrates a change in blood flow related to neuronal activity in a specific area of the brain.The largest growth area for radiology using biomarkers for surrogate endpoints has clearly been in oncology. In multiple trials of therapeutic agents for cancer, sponsors have used a decrease in tumor size, seen using various imaging modalities (including CT and MRI), as the surrogate endpoint in evaluating the response of solid tumors.5 Other growing areas include the application of surrogate endpoints in research involving the central nervous system and the musculoskeletal system. For example, in the evaluation of treatments for Alzheimer disease, classic endpoints have been based on clinical criteria and neuropathology. Due to the natural course of the disease, it can take large amounts of time to see the effects of therapeutic agents. Since the advent of biomarkers, imaging biomarkers, and surrogate endpoints, researchers have been using genetic markers, such as apolipoprotein E4; cerebrospinal-fluid markers, such as beta-amyloid and tau protein; and imaging biomarkers, such as hippocampus/entorhinal-cortex volumes and structural connectivity, evaluated using diffusion tensor imaging and functional MRI (see figure).6Trial Methodology Involvement in clinical trials represents an opportunity for growth for a high-quality imaging center or group. While typically, an imaging contract-research organization (CRO) handles the specific protocol design and implementation for sponsors, there is an opportunity for imaging centers to participate. A typical imaging based trial involves three components: the protocol, the imaging sites, and the imaging CRO. Protocol: An imaging protocol is designed that pertains to (and details) the imaging biomarker being used to study the outcome of a therapy. This protocol defines an outline that standardizes the method in which the imaging is performed, as well as the format in which images are stored, transmitted, and reviewed. Imaging sites: Typically, an imaging CRO will seek imaging centers or groups to assume responsibility for acquiring the images. The imaging site is responsible for scheduling patients, ensuring that the protocol is followed, and using the predefined techniques for transmission of the images to the imaging CRO. It is imperative for the imaging site to provide high-quality images and for it to be able to follow the protocol, as incorrectly acquired data are discarded from the trial. In small trials—where a single patient can alter the statistical power of the methodology—it is crucial for there to be no mistakes. Imaging CRO: Once the images have been collected, it is the responsibility of the imaging CRO to ensure that the data are evaluated using the methodology outlined initially by the protocol. Typically, images are reviewed in a blinded fashion by radiologists from the imaging CRO. There can be as few as two reviewers (with a third adjudicating reviewer), or there can be multiple reviewers at multiple sites. As the provision of medical imaging evolves, in the new era of decreasing reimbursement, many imaging centers are seeking alternative revenue streams. If an imaging center or group provides high-quality, reproducible imaging that can support complex protocols and has radiologists who can both implement and test these protocols, clinical-trial imaging can be a good avenue for diversifying revenue streams. Amit Mehta, MD, FRCP, is a radiologist with South Texas Radiology Group, San Antonio, and consulting radiologist and therapeutic lead for a contract-research organization; dramitmehta@IntrinsicCRO.com.
Figure. Functional MRI demonstrates a change in blood flow related to neuronal activity in a specific area of the brain.