Introduction
Imaging has changed beyond recognition in the last forty years, with new imaging techniques, new applications and new opportunities for clinical trials. Planar X-rays and ultrasound have been joined by digital subtraction angiography (DSA), X-ray CT, single photon emission computed tomography (SPECT), positron emission tomography (PET), optical techniques and magnetic resonance imaging (MRI).
A typical commercial clinical trial utilizing imaging will involve the sponsoring pharmaceutical company, a clinical trials CRO, one or more Imaging Core Labs (depending on the number of imaging modalities being used), possibly a specialist image analysis company or university lab, clinical sites and imaging sites. Imaging clinical trials typically involve 10-1000 patients scanned at 1-100 imaging centers and last from one to five years.
In this article I will draw on my own area of work, the application of MRI to the brain in psychiatry and neurology, particularly neurodegenerative diseases such as Alzheimer’s disease. Here MRI can be used in a number of ways, which are typical of the wider use of imaging in clinical trials. MRI can be used as part of the screening process to rule out any unwanted pathology such as tumors, vascular dementia or stroke in the case of Alzheimer’s disease. MRI can be used to detect any unwanted side effects of a new therapy such as micro-hemorrhages or edema. Semi-quantitative scales can be used by an experienced observer to describe key clinical features of the disease, such as the degree of medial temporal lobe atrophy in Alzheimer’s disease. Finally, MRI can be used as a quantitative technique, using expert observers or sophisticated image analysis techniques to determine volumes of key areas of the anatomy, such as the hippocampus and entorhinal cortex, which change early in Alzheimer’s disease, or the ventricles or whole brain volume. Other imaging modalities such as PET can alternatively measure function in a qualitative or quantitative manner. Typically, MRI will be used to determine changes in these measures over time and assess the impact of a new therapy on these measures during a clinical trial.
The Challenges of Imaging in Clinical Trials – Why an Imaging Core Lab is Needed
Clinical trials using imaging can be challenging to perform. The vast majority of modern imaging equipment is designed for use in clinical radiology departments to provide quantitative images for interpretation by clinical radiologists. Research and clinical trial work makes up approximately 1% by volume of hospital imaging workload and although a factor in the choice of equipment for some university hospitals and research units, clinical considerations typically dominate when buying radiology equipment. Manufacturers of imaging equipment aim to differentiate themselves from their competitors which makes standardization of imaging across different manufacturer’s imaging systems difficult. They are providing a scanner for quantitative imaging – not a highly calibrated scientific instrument. Imaging can be very complicated, particularly where quantitative measures are required and so experts in imaging are needed in order to make sense of the technology and to produce reliable results at a cost-effective price in an acceptable time scale.
Imaging clinical trials require good working relationships between clinical sites and their local imaging site. For a relatively rare and expensive imaging modality such as PET, several clinical sites may refer to a single imaging site, while for more prevalent lower-cost modalities such as planar X-ray or ultrasound, imaging may take place using several different pieces of imaging equipment at a single site. When choosing sites for a clinical trial, it is important that both the imaging sites and clinical sites are carefully assessed. The requirements for clinical diagnostic imaging are very different from those for clinical trials. Typically, diagnostic imaging requires radiographers or technologists to be able to modify protocols on the fly to acquire clinically acceptable images in as short a time scale as possible. 
This is in contrast to the strict requirements for consistency of imaging protocols needed for clinical trial work. The quality of images that a radiologist may be able to report from are often very different from the higher-quality images typically required for quantitative volumetric measurement, for example. Hospitals are rightly concerned about patient confidentiality, but hospital IT policies can run counter to the needs of clinical trials where rapid transfer of anonymized data to the global lab is important. Finally, imaging equipment is periodically upgraded or modified, up to several times a year, which again can run counter to the needs of clinical trials which require stability of performance. For example, manufacturers of MRI equipment compete with each other for their scanners to run faster than other manufacturers. This can mean that an imaging protocol that works well on one software level is altered without the user’s knowledge when a software upgrade is installed, causing important changes in key imaging parameters and negatively impacting on the clinical trial. Routine system maintenance, or replacement of key components can also cause unexpected change in the performance of the imaging system and this needs to be assessed and controlled for during clinical trials.
What Services do Imaging Core Labs Provide?
Imaging Core Labs provide a number of services in global clinical trials. A first important step is providing advice on the choice of imaging modality for the disease and therapy in question. Different imaging modalities can provide complementary information for either a primary or secondary endpoint, with the accuracy and precision of the measures varying. Imaging modalities also have different advantages and disadvantages. For example, MRI provides excellent soft tissue contrast, avoids non-ionizing radiation and can be repeated many times. However, patients with some metallic implants such as most cardiac pacemakers cannot be scanned due to safety concerns and a small proportion of patients with claustrophobia or who are very obese will be unable to tolerate a scan.
Once the choice of imaging modality or modalities has been made, the Imaging Core Lab will often be involved in the design of the imaging component of the clinical trial, including the modality-specific quality control and quality assurance measures required. They will typically produce a separate imaging manual for the trial, paper or electronic forms for the imaging component of the trial, standard operating procedures and other key documents. Depending on the complexity and size of the clinical trial they may help with the setup of the imaging sites either by attending the sites in person, or by the production of training materials for the sites. Secure image transfer which respects the confidentiality of participants is critical and the Imaging Core Lab will typically specify the details of the image transfer, normally providing a range of options to allow for differing national and hospital requirements. Although there is a move towards electronic transfer of images via secure internet connections, it is important to provide for transfer by CD or/DVD too. Images can be stored on a picture archiving and communication system (PACS system) or within a specific imaging database, which the Imaging Core Lab will be able to either specify, or provide.
Imaging Core Labs typically specify the responsibilities of the imaging sites and any other site such as a university or hospital department carrying out image analysis for a trial. They normally provide central quality control and quality assurance of imaging. This may involve specification for testing of imaging equipment at the imaging sites and assessment of the patient images once they have been acquired. They will continually feedback to imaging sites on their performance and provide advice on improving and maintaining quality. Depending on the clinical trial in question they may also provide central reading of images by a radiologist or other trained expert and often carry out the image analysis, whether this is assessment according to a semi-quantitative scale, exclusion of negative side effects of the therapy, or quantitative volume measurement.
Imaging sites will normally be responsible for imaging examination of patients to standards specified by the Imaging Core Lab, imaging of test objects known as phantoms, and shipment of data via an approved study route to the imaging data coordination center. They will perform local imaging quality control to check that the imaging QC criteria for the clinical trial has been fulfilled, for example ensuring that patient movement has been minimized and that image signal intensity is uniform.
What are the Benefits of Imaging Core Labs in Phase II and Phase III Clinical Trials?
Imaging Core Labs can bring many benefits to a clinical trial. They allow pharma to concentrate on its core business, bring breadth and depth of clinical trials imaging experience to the table, and allow pharma to expand or contract access to imaging expertise in response to demand. Potentially, the use of Imaging Core Labs can provide cost savings, due to both economies of scale and the provision of imaging expertise only when needed. Finally, through their knowledge of the industry, clinical and academia facilities can also help in locating appropriate imaging sites for clinical trials.
How can Imaging Core Labs help to Overcome Problems in Using Centralized Imaging Services, for example, the Difficulty in getting Appropriate Data from the Investigating Sites?
One very clear role here is with pre-assessment of imaging sites. Radiology departments and clinical departments can vary greatly in their expertise and experience in clinical trials. A renowned clinical department that has a good reputation for reliable recruitment to clinical trials may be located in a hospital or university with an imaging department which is less experienced in clinical trial work, or has few imaging slots available for clinical trial work due to a very busy radiology workload from other clinical departments. 
Careful recruitment of high-performing clinical departments without an equally strong focus on the pre-assessment of the imaging site can be particularly problematic. It is difficult to underestimate how important training is for imaging sites, fully covering the range of imaging staff involved in clinical trial work including radiologists, radiographers/technologists, radiochemists and imaging physicists depending on the imaging modality for example. Assessment of imaging sites must include their capacity to scan clinical trial patients in addition to their existing clinical workload and the stability of their imaging equipment, for example, avoiding any hardware or software upgrades during the timescale of the clinical trial. Approval for an imaging site to commence work on a clinical trial should only be given once quality control, training and a dry run of the study has been completed. Finally, as elsewhere, open lines of communication are important for the success of the imaging component of clinical trials.
What are the Best Strategies for Finding an Imaging Lab that can Transfer Imaging Data Quickly and Effectively?
While some aspects of imaging in clinical trials are specialized, generic competencies are also key for choosing an Imaging Core Lab. Contract management, an intelligent client function and a payment structure which encourages and rewards the key imaging outcomes are all critical function considerations when choosing a good Imaging Core Lab. A good Imaging Core Lab will be proactive and characterized by excellent good communication skills. They will have a strong focus on a pre-assessment strategy which focuses on both the imaging sites as well as the clinical sites. A focus on clinical site choice at the expense of the choice of imaging site can prove very costly further down the line. When assessing potential Imaging Core Labs it is important to consider the Core Lab organizational structure, their training, quality management, information security management, risk management, development life cycle, change management, backup and recovery, access management, incident management and business continuity/disaster recovery. While all of these are critical, experience and good management should never be underestimated, and the procurement strategy for an Imaging Core Lab should always carefully balance quality and cost, since neither on their own is acceptable.
Andy Simmons is a Reader in Neuroimaging and Consultant Clinical Scientist (Medical Physics) and has published more than 140 papers on both methodology development and the application of MRI to clinical cohorts. He is the imaging lead for the multi-centre European AddNeuroMed Alzheimer study and the imaging coordinator for the NIHR Biomedical Research Centre for Mental Health. He has designed the imaging component of a number of investigator-led and commercial clinical trials and has worked as a participant in a large number of imaging clinical trials