Clinical Translation of Nanomedicine.

1. Introduction Nanomedicine, the application of nanotechnology to health and medicine, is a relatively new area of interdisciplinary science. The field involves a wide range of scientific disciplines, including physics, chemistry, engineering, biology, and medical science. The term nanomedicine can be traced back to the late 1990s and first appeared in research publications in the year 2000.1 Despite the wide adoption of the term nanomedicine, its definition varies among experts in this area.2 Some define nanomedicine broadly as any science that involves matters that are nanoscale. For example, the European Science Foundation in 2004 defined nanomedicine as “the science and technology of diagnosing, treating, and preventing disease and traumatic injury, of relieving pain, and of preserving and improving human health, using molecular tools and molecular knowledge of the human body”.2 While such a broad definition is all encompassing, it can be confusing. For example, such a definition would include traditional scientific fields such as molecular biology as part of nanomedicine, because molecules such as nucleic acids and proteins are also nanoscale materials. However, scientists have been studying these molecules decades before the term nanomedicine was even coined, and their research generally does not take advantage of unique properties that only exist for nanomaterials. A narrower definition of nanomedcine is the application of nanoscale material in medicine that takes advantage of the nanomaterial's unique properties.1 This Review will adopt this narrower definition in our discussion of the clinical translation of nanomedicine. Nanomedicine has made a rapid and broad impact on healthcare. Despite being only several decades old, research in nanomedicine has already led to the development of a wide range of products including therapeutics, diagnostic imaging agents, in vitro diagnostics, and medical devices. There are more than 200 nanomedicine products that have been either approved or are under clinical investigation.3 On the other hand, successful clinical translation is a challenging process. It requires extensive preclinical research, carefully selected clinical indication, proper design of clinical trials, and the successful completion of these trials. Mistakes in clinical translation can be unforgiving. Unlike preclinical research where there are many if not unlimited chances of generating a successful study, a single failed clinical trial can doom a drug's translation. Hay et al. recently showed that the eventual success rate of approval for therapeutics entering phase I trial is only about 10%.4 Because of this sobering statistic, it is important for translational researchers to fully understand the clinical translation process and to develop a successful translation strategy in the early stages of research. As compared to diagnostics and devices, clinical translation of therapeutics is arguably the most challenging. The typical clinical translation path for a new drug starts with investigators generating robust preclinical data to demonstrate the safety and efficacy of the new drug to enable an investigational new drug (IND) application with the Food and Drug Administration (FDA).5 Once the FDA has approved the IND, the therapeutic will be evaluated in a first-in-human or a phase I clinical trial. The goal of such a study is to determine the safety profile and pharmacology of the drug. It will result in a dose and schedule for further clinical investigation, or the recommended phase 2 dose (RP2D). The typical phase I trial design used a “3 + 3” cohort expansion design.6 This design assumes toxicity increases with dose, and it aims to determine the dose level that has less than 1/3 chance of a dose-limiting toxicity (DLT).7 In general, such a trial starts with a low drug dose. If none of the three patients receiving this dose experiences a DLT, another three patients will be treated at the next higher dose level. If one of the three patients experiences a DLT, then three more patients will be treated at the same dose level. Dose escalation continues until two patients among a cohort of three to six patients experience DLT. The RP2D is the dose level just below this level. Dose escalation typically follows a modified Fibonacci sequence where dose increments decrease as the tested dose increases. Other types of phase I designs include the accelerated titration designs, Bayesian models-based designs, and many others.7 Each design has advantages and disadvantages, and investigators have to choose the design that best fits the therapeutic. The goal of a phase II clinical trial is to examine the effectiveness of a drug or treatment. Secondarily, it will acquire more data on the toxicity and tolerability of the therapeutic. Therapeutics will progress to phase III clinical investigation only if they can demonstrate efficacy in phase II. The designs of phase II trials are either single-arm trials or randomized trials.8 Single-arm trials are cheaper, require fewer patients, and are typically easier to accrue. However, the outcome is less reliable as there is no comparison/control arm, and data are more susceptible to bias. Data from randomized phase II trials are more predictive of phase III results. However, it requires more patients and can be more difficult to accrue. Randomized phase II trials do not replace phase III investigations. Although they are randomized, patients are generally stratified on the basis of very few variables, such as age, sex, and disease status in phase II trials to keep the accrual goal low. Randomized phase III trials stratify patients on the basis of a large number of variables, which leads to less bias and more robust data. Because of the stratifications, the sample size required for phase III investigation is much higher than that of randomized phase II trials. The goal of randomized phase III trials is to demonstrate that the investigational treatment is more effective than the “gold standard” treatment. In general, phase III data are required for FDA approval. However, in select cases where there are robust data and unmet clinical needs, conditional approval can be granted on the basis of phase II data or interim phase III data. The FDA has a range of programs to speed up the approval process, including accelerated approvals and the recent “break through therapy” designation.9 There is a “short-cut” to FDA approval for agents that are based on already approved drugs. This pathway is called the 505(b)(2) pathway. The process of timeline for 505(b)(2) is much more abbreviated when compared to a typical approval process. For nanomedicine, this pathway will typically require that the exact nanoparticle platform is already approved with another agent and the drug being delivered by the nanoparticle is also approved. Past examples of this include the approval of liposomal bupivacaine with the DepoFoam liposome platform. The FDA was granted the authority to regulate medical devices in 1976.10 The approval process for medical devices is very different from that of drugs. First, for devices that predate May 28, 1976, these devices can remain on the market without needing approval. For the devices entering the market after that date, they are classified into different classes (I, II, and III) on the basis of their risks (Table 1).10 Class I devices are of low risk and are generally exempt from premarket notification (referred to as 510(k)) and may even be exempt from compliance with the good manufacturing practice requirement. Class II devices typically will require 510(k) submission before marketing. Class III devices are subject to the most stringent regulatory controls. Their approval will require a premarket approval (PMA) application. The 510(k) pathway is for devices that can be compared to existing, legally marketed “predicate” devices. The new device needs to be shown to be at least as safe and as effective as the “predicate” device. For devices that do not have a “predicate” device with which to compare, they are classified as class III and will need PMA. PMA needs to include scientific evidence that the device is safe and effective for its intended use. Unlike therapeutics where approvals generally require large randomized studies, scientific evidence for devices can include randomized controlled trials, single-arm studies, well-documented case series, and reports of significant human experience. For new devices that pose significant potential risks, an investigational device exemption (IDE) application is required prior to clinical investigation. Overall, the approval process is much simpler for devices than for therapeutics. Table 1 Summary of the FDA Device Regulation Processa In this Review, we will examine preclinical evidence, chosen clinical path to translation, and clinical data of clinically approved nanomedicine products. We will also discuss the clinical data on nanomedicines that are under clinical investigation or failed clinical translation. Each of these clinical nanomedicine products has a unique clinical translation story. By examining this body of evidence, we aim to formulate important concepts that are keys to nanomedicine's clinical translation and to identify challenges. Such concepts will facilitate the translation of future nanomedicine products.