Husseini Manji suggested that from a regulatory perspective, moving a therapy forward that involves novel methods for crossing the blood–brain barrier (BBB) enters into uncharted territory. Thus, he said, it makes sense to seek the advice of regulators early in the process. Matthew Whittaker concurred, adding that face-to-face meetings between sponsors and the clinical review team at the Food and Drug Administration (FDA) are critical to answer questions associated with trial design.
Vikram Patel, deputy director of the Division of Applied Regulatory Science in the Center for Drug Evaluation and Research at FDA, raised four areas of concern with the technologies discussed: off-target activity, safety of long-term dosing for chronic diseases, immunogenicity, and unintended consequences. For example, he noted that off-target activity could be particularly troublesome when using the transferrin or insulin receptors since they are ubiquitous. To ensure the safety of chronic dosing, more data will be needed. With regard to immunogenicity, he recalled that although checkpoint inhibitors appeared to be safe in preclinical models, when they were introduced in the clinic, patients died as a result of high-level immune activation that caused excessive, systemic autoimmune responses. He suggested that new technologies, such as humanized mouse models, may enable prediction of this reaction. There may also be other unexpected consequences to therapy with these novel therapeutics, he said.
Opening the BBB with focused ultrasound or other technologies raises other issues of particular relevance to regulators, said Patel. As mentioned earlier, for example, it may be desirable to close the BBB as soon as possible to limit exposure of the brain to chemicals or toxins other than the intended therapeutic compounds. Patel said that for some treatments, strategies may also be needed to keep drugs in the brain longer so that dosing can be less frequent, possibly resulting in a reduction of toxicity. Specific issues related to preclinical and clinical phases of development are discussed in more detail below.
Matthew Whittaker said that he and his colleagues at FDA evaluate the nonclinical data to determine if a study should proceed, and then inform clinical reviewers on the review team of potential risks that may exist. With biologics, unexpected challenges arise, said Douglas Hunt. The job of a company’s regulatory affairs department is to try to project where things can go wrong, formulate questions that need answering to address specific challenges, and then look at the data to try to answer these questions and assess the risks and benefits, he said. Danica Stanimirovic noted that since Trojan horses are quite complex—usually at least bifunctional—and may combine antibodies, enzymes, and growth factors that will be delivered into the brain, evaluating safety may be particularly difficult.
Another potential complication that can affect both safety and efficacy of biologics is immunogenicity, said Eric Schaeffer. To avoid evoking strong immune reactions in preclinical models, different antibodies may be required for studies in rodents versus primates, he said. For example, human proteins may elicit a strong antibody response in rodents, which generally means that safety studies must be conducted in at least one rodent and one nonrodent species, said Whittaker. He said that sponsors address these kinds of issues in different ways with guidance from FDA. Patel noted that even with creative solutions and apparently adequate safety studies in animal models, there may still be surprises. Thus, sponsors must be prepared to carefully monitor safety in phase I studies, he said.
In addition to safety, Francesca Bosetti reminded workshop participants to also keep efficacy in mind. She said that preclinical, translational programs at the National Institute of Neurological Disorders and Stroke try to mimic as closely as possible the practices used in clinical trials, keeping in mind that outcome measures in clinical studies should reflect outcome measures that are used as end points in clinical trials. Hunt agreed, noting that in order to provide animal data on safety and efficacy relevant to humans, selecting the right model for preclinical studies requires not only that the therapeutic molecule crosses the BBB but also that it finds the correct receptor in the brain.
In terms of safety, the regulatory requirements for a drug intended for an orphan indication are no different than for any other drug, said Whittaker. Generally, this means that there must be an acceptable margin
between the observed adverse effect level in animals versus the maximum dose that the sponsor intends to give humans, he said, noting that this includes adverse effects in the brain as well as in other organ systems throughout the body.
Moving from preclinical to clinical development, E. Antonio Chiocca, Harvey Cushing Professor of Neurosurgery at Harvard Medical School, expressed concern about using animal models for safety assessments, since animal models may have limited reliability as tools for assessing human toxicity. However, Whittaker commented that for first-in-human trials the only safety data available may have come from animal studies. One approach some sponsors have taken, he said, is to integrate all available in vivo and in vitro data to establish the minimal anticipated biological effect level, followed by incremental dose increases (Muller et al., 2009). Patel added that while a sign of toxicity in animal models may not be relevant to humans, it cannot be ignored and must be followed up with modeling and other studies.
Another concern about the use of animal models was raised by Bosetti, who noted that while small animals present fewer ethical challenges and are cheaper and widely available, they may lack relevance in terms of diffusion, since it is much easier to reach all areas of a smaller brain compared to a larger one. Bosetti added that the acceptable safety threshold for a therapeutic agent will vary depending on the condition being treated. For example, patients may accept a lower safety threshold for an aggressive lethal brain cancer than they would for a long-term neurological disease, she said. Thus, when evaluating a treatment such as focused ultrasound, she suggested that developers and regulators may need to consider the potential for long-term cognitive effects. Subtle cognitive effects would be difficult to detect in nonclinical studies, said Whittaker and Patel. Hunt suggested that a registry would likely be needed for long-term follow-up.
William Potter, senior advisor to the director at the National Institute of Mental Health, raised another safety issue that could come up in phase I clinical studies. For setting doses, it may be necessary to expose trial participants to radiolabeled tracers that indicate the concentrations of drug reaching relevant areas of the brain. To determine the risks of that exposure, Deepa Rao said that preclinical toxicology studies in animals
would be the first step, and Patel added that in humans it would be important to use radioactive compounds with a relatively short half-life.
Once clearance is given by FDA to allow testing in humans, the risk–benefit ratio becomes the key point of discussion, said Patel. Hunt noted that the risk–benefit ratio is different for each indication. One challenge he cited in assessing risks and benefits for orphan indications is the appropriate clinical end point may be unclear. A bigger challenge in terms of conducting a clinical trial for an orphan indication is enrollment and how to deal with heterogeneity when the number of trial participants is small, he said.
Alexandra Golby said that she and her colleagues are in discussions with FDA to try to put together the first clinical trial of focused ultrasound with microbubbles to disrupt the BBB and deliver drugs. She added that this approach poses additional regulatory challenges since it combines a device, microbubbles, and an imaging agent, plus the drug that is being delivered. However, she noted that the devices (the device for thermal ablation and the microbubbles for imaging although neither has been approved for BBB opening) imaging agent, drugs, and microbubbles have already been approved for use. A regulatory package will also require substantial work to determine what volume of the brain can be safely targeted with this approach and whether repeated treatments can be delivered. And, most importantly, the right drugs are needed, she said. Therapeutic agents that have already been approved for other indications are being considered, Golby added.
While consideration of cost is not part of the FDA mandate, Hunt and others commented that costs are often of concern to people involved in drug development as well as to society as a whole. With the advent of more expensive technologies, he suggested that citizens will have to participate in decision making about which treatments should be subsidized. Patel suggested that one way to lower costs of clinical trials might be by developing and qualifying biomarkers.
This page intentionally left blank.