Optimize Your Imaging

The role of medical imaging shifts dramatically between early and late phases of drug development, from detecting early signals to being a surrogate endpoint that determines safety and efficacy of a new treatment. Early on, sponsors are searching for biomarkers that characterize the pharmacokinetic and pharmacodynamic profile of an investigational compound. Later, they are looking to define surrogate endpoints as substitutes for clinical endpoints that could take years to achieve.

PHOTOGRAPHY : MEDICALRF.COM, GETTY IMAGES

Making the transition from early stage imaging in small studies to late-stage, multicenter trials can be challenging for sponsors who have not formulated a development plan to incorporate imaging technologies well ahead of their potential use in pivotal Phase III trials. For these sponsors, bringing imaging into earlier phases helps to establish a continuum in the development of potential biomarkers into potential surrogates. It also enables sponsors to make faster and better informed go/no-go decisions about the future of investigational compounds.

This article describes the role of imaging in each phase of clinical trials and how planning and regulatory guidances can help stakeholders incorporate imaging into the process to improve outcomes. It also focuses on how imaging changes from its early purpose of identifying biomarkers in small, single site studies to its later function as a tool for surrogate endpoints in multicenter trials.

Early imaging

Much of the emphasis on imaging has traditionally occurred in later phase studies, using technologies such as magnetic resonance imaging (MRI) and multislice computerized tomography (CT). But with advances in methodologies such as positron emission tomography (PET), single-photon emission computed tomography (SPECT), and functional magnetic resonance imaging (fMRI), a gradual increase in the amount of early phase imaging is taking place.

The intent of imaging differs among Phases I, II, and III (see Table 1), as it moves from serving mostly an exploratory function in early studies that are not well-controlled to large, multicenter, well-controlled clinical trials (see Figure 1) with defined endpoints and outcomes.

At the beginning, the goal of imaging is to find prebiomarkers, which are signals, and the meaning of the initial results may not be clear. The work is typically conducted at a single site or at a handful of investigative sites, mostly at academic institutions. The same investigators who enroll and manage the research patients also perform the imaging and image analysis, so by their very nature, the studies are biased. As the data generated are not FDA-bound, regulatory rigor is not a prerequisite.

Imaging in Clinical Trials

To start the process of incorporating early imaging into a clinical development plan, stakeholders might focus on maximizing its value. If an imaging modality or compound seems to hold promise, they might consider using the FDA's exploratory Investigational New Drug (IND) process, an underutilized tool that offers the chance to competitively analyze multiple imaging compounds or modalities prior to routine first-in-human studies.

Figure 1. A look at imaging and its purpose throughout the drug lifecycle, from early to late phase studies.

The exploratory IND is designed to reduce the threshold on the amount of pharmacologic and toxicologic testing on drugs and biologics needed before opening the IND and is a developmental pathway that takes technology in a direction that regulators should find acceptable if widely used. It can speed the process of finding biomarkers that may eventually emerge as imaging surrogates in subsequent Phase IIb and Phase III studies. Generally, exploratory IND studies involve very limited human exposure and have no therapeutic or diagnostic intent, but can serve a number of useful purposes, such as providing information on pharmacokinetic and pharmacodynamic properties as well as tissue target specificity. The FDA issued an exploratory IND guidance in 2006 with a goal of making drug and biologics development more efficient by outlining ways to expedite Phase I clinical trials through approaches meant to distinguish promising from doubtful candidates.

Applying Imaging to the Fast Track Process

The guidance indicates that by opening a single exploratory IND, companies can perform a batch assessment of an entire portfolio of imaging compounds simultaneously or sequentially, via whole-body biodistribution studies. Sponsors with a small portfolio of candidate products can also avail themselves of facilitating elements in the guidance and implement them directly under a standard IND.

Establishing Validity in Imaging

Later phase goals

As development continues, research moves toward developing biomarkers that help researchers learn about the safety of a compound; stratification of subjects, which refers to how various homogenous groups react to a compound; biodistribution, which determines how a compound travels within a human subject; and proof of concept. Using these data, investigators look for a correlation between an imaging outcome, such as brain atrophy, and a clinical outcome, such as worsening dementia. Each biomarker can also be evaluated in early dose ranging studies instead of waiting until Phase II to begin this work. These studies may take place in a small number of sites, and efficacy is not yet confirmed.

By Phase IIb and III, imaging serves to confirm efficacy. Imaging surrogates become useful substitutes for clinical endpoints, the studies are scientifically exacting, and imaging practices are harmonized across multiple sites. Surrogates offer significant advantages over traditional clinical endpoints, which have considerable variability based on a host of difficult to control factors, such as time points that can take years to achieve, and metabolic processes among subjects. By comparison, surrogate endpoints, once fully validated, can be a more precise, objective measure and provide ample statistical power using fewer subjects. One example of an imaging surrogate endpoint is progression-free survival as measured in lung cancer trials by CT or MRI scans that monitor changes in tumor size. Much time can be trimmed from development timelines when an imaging surrogate is used in oncology trials, depending on the patient population, therapy, and section of the tumor being targeted. As a result, the surrogate endpoint can help speed important therapies to market at lower costs.

Validating the surrogate endpoint

For a surrogate endpoint to be accepted by a regulatory agency, it must be validated under its own investigational new drug (IND) application. It also needs to reliably predict the overall effect of an intervention on the clinical outcome and detect what is being claimed. Regulatory input is needed to evaluate whether study protocols are taking a suitable approach and adhering to a principle known as "standard of truth." This principle refers to the gold standard outcome for the particular disease being studied.

In oncology trials, for example, survival is usually the standard of truth. In multiple sclerosis studies, it is most often a reduction in disease-related events such as paralysis and speech impairment. For certain therapeutic areas, investigational treatments tested via a surrogate endpoint tend to qualify for accelerated approval from FDA as defined by the FDA Modernization Act, contingent upon the sponsor's commitment to conducting a continuation study to confirm results against the standard of truth (see sidebar below).

In October 2008, three pharmaceutical companies met with FDA's Peripheral and Central Nervous System Drugs Advisory Committee to present imaging agents they had in development for diagnosing Alzheimer's disease (AD) and to discuss their plans for Phase III trials.

The Advisory Committee's purpose was to address the clinical development of radionuclide imaging products for the detection of amyloid to assist in the diagnosis of AD and to tackle the issue of selecting the appropriate reference, using the standard of truth guideline. The three companies have agents in the pipeline that are designed for use with PET scans to detect abnormal amyloid plaques in the brain. It is widely believed, though unproven, that amyloid causes AD. Currently, the presence of amyloid plaque in the post-mortem brain provides the only definitive diagnosis.

Company 1 is developing an amyloid imaging agent and proposed a comparison of brain images taken from live subjects with terminal conditions not related to dementia against demented and nondemented subjects at autopsy. The demented subjects had various stages of AD, mild cognitive impairment (MCI), and other causes of cognitive impairment. The primary endpoint is positive predictive value of low amyloid burden to rule out AD. The standard of truth relates to pathology detection of amyloid.

Company 2 sought to prove its imaging tracer used in conjunction with a PET scan could accurately detect amyloid beta plaque in the brain, and in the absence of the imaging uptake, could exclude a diagnosis of AD when compared to clinical data. And company 3 looked to assess its investigative imaging agent in patients with AD, patients with mild cognitive impairment, and in healthy age-matched volunteers. It proposed a similar amyloid imaging agent could serve as the standard of truth as a surrogate for amyloid.

The Advisory Committee supported the first company's correlation of the amyloid imaging agent with actual autopsy data as the standard of truth, and it is currently moving into Phase III. The other two companies' plans were deemed less preferable by the Advisory Committee when compared to pathology, the designated standard of truth for AD.

Letter of the law

An FDA guidance on imaging states that medical imaging agents are governed by the same regulations as other drugs, and recommends that investigations establish the validity and reliability of imaging agents (see the sidebar above). Trials must prove that any site can take the requisite image and generate equivalent results. The imaging guidance suggests that imaging surrogates be validated as representing the standard of truth for the particular disease being studied.

A companion imaging guidance encourages stakeholders to consider the method of image analysis when attempting to validate surrogate endpoints. The guidance suggests that imaging, reading, and assessment of images—particularly in Phase III—should be performed as a fully blinded image evaluation or as an image evaluation blinded to outcome by independent readers as the principal image evaluation for demonstration of efficacy. This is a transparent process that can be documented with an audit trail, and is in keeping with the spirit of the guidance. Processes that are not transparent may be problematic for regulatory agencies.

For example, if a variety of imaging agents are bundled together within an IND, and results are interpreted using a fully automated image analysis tool, the approach might be scientifically valid but may not meet regulatory requirements for transparency. Specifically, regulatory bodies may not consider the image analysis algorithm to be validated, raising concerns about the opportunity to modify images. A better route would be to pursue a separate IND for each imaging agent, which is later tested using independent reviewers in a transparent process.

Challenges along the way

In moving from early to late phase trials, sponsors face numerous challenges as they anticipate the potential of an imaging biomarker to mature into a validated surrogate. One of the key challenges is creating distance between the key stakeholders who have invested heavily, both financially and emotionally, in the imaging biomarker from the beginning, and its further study in later phase trials.

Generally, it is a difficult jump from the cloistered world of first-in-man trials piloted by a few key opinion leaders (KOLs) to a multicenter study with strict standards for obtaining and analyzing images. With ties to their early contributions, it may be hard for KOLs to accept their diminished role from top leader to just one investigator among many. It is important for sponsors to focus on the next stage of proving reproducibility of results, as the focus moves beyond the single center study to diverse, multicenter settings.

By the time Phase IIb studies are in the planning stages, sponsors should feel confident they have a biomarker that would make a strong candidate for imaging surrogacy. The trial can, therefore, mirror the look of a Phase III trial in terms of statistical power, harmonized imaging, and bias control via independent, unvested image readers. If successful, data from the trial might be submitted in support of product registration to regulatory agencies.

Unfortunately, the reality is that some study sponsors tend to simply perpetuate how biomarkers are used in Phase IIa trials—studied in a limited number of centers in studies that are inadequately controlled. If, however, sponsors tap the expertise of an imaging core lab at the protocol design stage, its experience in planning ahead can help sponsors and investigators design the Phase IIb trial with all the rigor of a Phase III. Bringing this dimension into the Phase IIb study can yield key insights to stakeholders vested with deciding whether to push ahead with a costly Phase III program or kill a less than promising compound.

There are other challenges in bridging the gap between the imaging biomarker and its development as a surrogate in later phase trials. Specifically, a promising imaging modality that is linked to a devalued therapeutic may be dropped from the development pipeline. The value of the imaging surrogate could, however, be saved if a new IND is constructed for it, separating out work on the product under the prior IND. But the more likely scenario is that the investigative diagnostic will lose visibility and study momentum.

Plan early and wisely

Detecting signals in early trials is different than using a validated imaging surrogate in late phase studies. The quickest, least costly path to approval requires a development plan early on that anticipates imaging's transition from a biomarker in small, KOL-managed biased studies to an FDA-approved surrogate in multicenter, bias-controlled trials.

Sponsors working from the premise that they have a commercially viable imaging methodology should consider opportunities presented by exploratory INDs to quickly triage drug candidates before investing in their development. Using this process as a foundation, they can determine the pharmacokinetic and pharmacodynamic properties of compounds from the beginning to identify and advance promising candidates while stopping less hopeful ones.

As biomarkers are identified that indicate safety and efficacy, they can play a role in later trials designed to evaluate those biomarkers as imaging surrogates. Potential surrogates must be validated through the IND process if regulatory agencies are to consider them as acceptable alternatives to traditional endpoints. By using surrogate endpoints in Phase IIb studies as opposed to starting them in Phase III, the timeline is compressed, fewer subjects are needed, and sponsors are in a better position to make critical decisions about forging ahead with costly Phase III studies.

With greater use of imaging, there is the overall potential to improve decision-making to accelerate all phases of the clinical development process.

References

1. Food and Drug Administration, Guidance for Industry, Investigators, and Reviewers, Exploratory IND Studies, (FDA, Rockville, MD, 2006).

2. D. Jacobson-Kram, Preclinical Safety Data for "First in Human" (FIH) Clinical Trials in Healthy Volunteer Subjects, Oncology Advisory Committee, FDA, March 13, 2006.

3. G. Mills, "The Exploratory IND," Journal of Nuclear Medicine, 49, 45N-47N (2008).

4. S.D. Curran, A.U. Muellner, L.H. Schwartz, "Imaging Response Assessment in Oncology," Cancer Imaging, 6, S126-130 (October 2006).

5. Food and Drug Administration Modernization Act of 1997.

6. Food and Drug Administration Advisory Committee Provides Clear Path Forward for Development of Amyloid Imaging Agents For Alzheimer's Disease, press release, October 23, 2008.

7. Food and Drug Administration, Guidance for Industry: Developing Medical Imaging Drug and Biological Products, Part 2 (FDA, Rockville, MD, 2004).

8. U.S. Department of Health and Human Services, Food and Drug Administration, Guidance for Industry, Developing Medical Imaging Drug and Biological Products, Part 3: Design, Analysis and Interpretation of Clinical Studies (FDA, Rockville, MD, 2004).

James Paskavitz, MD, is director of exploratory imaging and medical director of CNS imaging at Perceptive Informatics, 2 Federal Street, Billerica, MA 01821, email: james.paskavitz@perceptive.com