Proposal for a “phase II” multicenter trial model for preclinical new antiepilepsy therapy development


Address correspondence to Terence O'Brien, The Department of Medicine, The Royal Melbourne Hospital, Royal Parade, Parkville, 3050, Victoria, Australia. E-mail: or Michele Simonato, Neuroscience Center, University of Ferrara, via Fossato di Mortara 17-19, 44100 Ferrara, Italy. E-mail:


There is a pressing need to address the current major gaps in epilepsy treatment, in particular drug-resistant epilepsy, antiepileptogenic therapies, and comorbidities. A major concern in the development of new therapies is that current preclinical testing is not sufficiently predictive for clinical efficacy. Methodologic limitations of current preclinical paradigms may partly account for this discrepancy. Here we propose and discuss a strategy for implementing a “phase II” multicenter preclinical drug trial model based on clinical phase II/III studies designed to generate more rigorous preclinical data for efficacy. The goal is to improve the evidence resulting from preclinical studies for investigational new drugs that have shown strong promise in initial preclinical “phase I” studies. This should reduce the risk for expensive clinical studies in epilepsy and therefore increase the appeal for funders (industry and government) to invest in their clinical development.

Despite the licensing and marketing over the past three decades of multiple new medications to treat patients with epilepsy, there has been little effect on the prevalence of 30% of patients with drug-resistant epilepsy (Sarter & Tricklebank, 2012; Galanopoulou et al., 2013). Furthermore, there are currently no medical treatments available that are disease modifying (i.e., that prevent the development of epilepsy following a brain insult or mitigate it once it is established) or are able to address the neuropsychiatric comorbidities that commonly accompany epilepsy and that significantly affect quality of life. Despite these, and other, considerable unmet medical needs for patients with epilepsy, the pharmaceutical industry and other funders have become increasingly reluctant to invest in developing potential new antiepileptic therapies (Galanopoulou et al., 2012).

A major contributor to this reluctance is an increasing concern that results from preclinical target validation or drug testing studies are often not replicable in independent studies, including in-house testing by industry (Prinz et al., 2011; Landis et al., 2012). Furthermore, there is a reasonable concern that the results of preclinical studies often fail to translate into positive results in clinical trials (Prinz et al., 2011; Kimmelman & Anderson, 2012; Landis et al., 2012). The failure of clinical trials is expensive and provides a major disincentive for further development in the field. Although these issues are not specific to epilepsy, the failure of the many new antiepileptic drugs that have been introduced to demonstrate increased efficacy over the established medications is an important factor that negatively influences the development of new antiepileptic therapies by industry. A major advance would be to demonstrate that preclinical drug development is predictive of superior clinical efficacy against drug-resistant seizures, or efficacy against epileptogenesis or comorbidities. To date, however, such promise has not been realized.

There are multiple reasons for the failure of replication and translation of positive preclinical results; publication of inadequately powered studies and other biases, as well as lack of study design rigor, likely play a role (Prinz et al., 2011; Landis et al., 2012). In addition, preclinical studies rarely compare the new therapy against established therapies in the same experiment to demonstrate differential efficacy.

The International League Against Epilepsy/American Epilepsy Society (ILAE/AES) Working Groups joint meeting to optimize preclinical epilepsy therapy discovery proposed a potential solution to address this problem by implementing a “phase II” multicenter preclinical trial paradigm modeled on the methodology used for double-blinded multicenter clinical trials. Phase II preclinical trials would be adequately powered and executed to detect clinically relevant differences in efficacy, involving a larger number of animals than can be achieved in single laboratory studies. This approach would reduce biases related to individual laboratory practices and conditions, implement rigorous blinding and statistical design, and incorporate independent monitoring of data collection and analysis.

Phase I: Proof-of-Concept Preclinical Studies

For the purposes of differentiating our proposed “phase II” multicenter trials, we have termed the current standard preclinical studies as “phase I.” This is not meant to imply that all therapies need necessarily to be tested in both phase I and phase II preclinical studies sequentially, as discussed later in this article. Phase I studies provide initial proof-of-concept data, generated by an investigator, group of investigators, or a company, that a potential therapy (pharmacologic or nonpharmacologic) may possess efficacy against drug-resistant seizures, epileptogenesis, disease modification, or comorbidities. These studies are usually undertaken in a relatively small number of animals from one (or a few) laboratories using the particular animal models in which they have experience and expertise. We believe that this should remain as the mainstay of drug discovery, but propose that phase II preclinical testing will provide an option to improve the evidence base for efficacy prior to proceeding to expensive clinical trials with a potential therapy or further, extensive drug discovery efforts on a mechanistic pathway.

Therapies potentially eligible to enter phase II preclinical testing will emerge from studies from either traditional academic laboratory based investigator-initiated research or from industry.

Selection of Therapies for Phase II Preclinical Studies

The criteria for therapies to enter phase II preclinical studies will need to accommodate the fact that potentially eligible candidates could come from a variety of sources at different phases of preclinical development. For example, therapies could either represent a proof-of-concept intervention or a “true” preclinical drug candidate that has passed the testing required to be registered as an investigational new drug (IND) (Fig. 1).

Figure 1.

An outline the current developmental pipeline for the development of new drugs. Candidates for phase II preclinical studies could be drawn from different stages of this pipeline, from early “proof-of-concept” therapies to later “true” preclinical drug candidate that has potential ultimately to be a therapeutic drug in clinical practice.

We propose that rigid criteria are not imposed to select suitable candidates for phase II preclinical testing. The rationale for the phase II testing is that therapies demonstrating efficacy in phase II preclinical trials will be more likely to be efficacious in subsequent clinical trials. Therefore, the fundamental criteria is that a therapy shows sufficient promise in phase I preclinical studies to warrant further investment in drug development, but “derisking” of the clinical studies is desired.

Usefulness of phase II preclinical studies to fill critical knowledge gaps between early phase I preclinical studies and clinical trials

Several factors should be taken into consideration when selecting therapies for phase II preclinical trials:

  1. Level of evidence from phase I preclinical studies: Although it is ideal that phase I preclinical studies have demonstrated efficacy in more than one model and more than one species and one laboratory, it is important to take into account the number of models that are available for the particular syndrome for which the new treatment is a target. For example, if only one model is available, positive data in that model may suffice. Moreover, if a therapy proved effective in more than one model and/or species, then its value is increased and the value of a preclinical phase II stage is potentially reduced. Therefore, in some circumstances, phase I evidence may be considered sufficient for moving directly to clinical studies. Phase II preclinical studies will be more costly, and time and resource intensive than traditional “phase I” preclinical studies (although considerably less than clinical efficacy studies), and there will be limitations in the capacity to test many different therapies. Accurate selection of candidates and avoidance of useless repetitions will be important.
  2. Potential advantages over currently employed therapies: Ideally, comparative efficacy and tolerability studies against reference therapies are needed for this. However, this kind of experiments requires larger numbers of animals than most laboratories are prepared to handle. Therefore, comparative efficacy would be an important role of phase II preclinical trials.
  3. Pharmacokinetics (PK) and toxicology: PK and toxicology data are not strictly needed for access to phase II preclinical studies. Detailed PK and toxicology studies are not generally made early in discovery (see Galanopoulou et al., 2013, in this Supplement). They could be additional services provided by institutions to help in the development of promising therapies, but this needs to be partnered with the pharmaceutical industry if the compound shows promise. The pharmaceutical industry has the best resources for final PK and toxicology studies.
  4. “Clinical translatability” of the phase I preclinical data: There is often a significant lack of parallels between the models and the clinical condition that they are expected to model (Sarter & Tricklebank, 2012). Phase II preclinical studies should employ the best, most clinically relevant models available for the epilepsy syndrome being targeted by the potential new therapy. Failed human trials often reflect insufficient understanding of the conditions necessary for the candidate therapy to work and of the relationship between pharmacodynamics (PD) and PK of such candidates. Phase II preclinical studies should address these aspects.

Who will decide which therapies are selected for phase II preclinical studies?

This will depend, at least in part, on the source of the funding for the study. If the founder is industry, the decision will be made by the industry itself. If public funds are involved, this should be determined by an independent Scientific Advisory Committee made up of experts from the field, including basic scientists, clinical trialists, representatives of industry, regulatory agencies, and funding agencies. The Committee should consider the current scientific literature to identify and prioritize candidates for phase II preclinical study. It should be explored whether a “Cochrane type” collaborative group could be formed to support this work.

Outline of the Model for Phase II Multicenter Preclinical Studies

These are proposed to be multicenter studies designed along the lines of phase II/III clinical studies: involving a number of laboratories using the same models and methods, a relatively large number of animals in total (divided between a number of centers), standardized methods and end points, rigorous statistical and sample size calculations, rigorous blinding, and independent data monitoring and analysis (Landis et al., 2012).

Efficacy testing

Assessment of the efficacy of the potential new therapy against clinical relevant end points where there is currently a treatment gap (i.e., drug-resistant seizures, antiepileptogenesis, disease modification, or comorbidities) is the primary purpose of these phase II preclinical studies.

The studies would be designed as multicenter, double-blinded, randomized controlled trials. The studies should compare the new therapy with an inactive control (i.e., vehicle), and ideally also with at least one appropriate, established antiepileptic drug currently in clinical practice so that evidence could be sought for incremental efficacy of the new therapy being developed. Ideally, 5–20 laboratories should participate in a particular phase II preclinical trial, and 2–5 different models should be used. The choice of the models would be on the basis of the presence of readily quantifiable efficacy end points that relate to specific syndromes and gaps in clinical care (see other articles in this Supplement). However, in general it would be envisaged that these would be “true epilepsy” models that exhibit spontaneous recurrent seizures. The number of animals per study would be based on power calculations from phase I preclinical studies performed by expert biomedical statisticians with experience in clinical trial design. The number of subjects required to be studied is likely to be less than that in clinical studies because of the lower variability (i.e., noise), and more suitable study conditions and end point assessments in experimental models than in clinical studies (e.g., measuring seizures with continuous electroencephalography (EEG) recordings rather than patient-reported seizures). The studies would involve standardized methodology, rigorous blinding, and well-defined criteria for end points (i.e., EEG, behavior, and so on).

There would be a central coordinating site for each study, which should be independent from the data collections sites. Data collected at individual sites could be analyzed (blinded to treatment arm) at least in part by the local sites for relevant end points (e.g., seizure quantification, neurobehavioral testing). The raw and/or analyzed data would be transmitted to the central coordinating site for any higher level analysis and pooling with data from other sites for the final data analysis. There would also be independent monitoring of data collection and analysis, as currently occurs with multicenter clinical trials.

Toxicity testing and pharmacokinetic (PK) data

Although not the primary purpose of phase II studies, ideally (even not necessarily) some limited drug tolerability (see Galanopoulou et al., 2013, in this Supplement) and PK data would be collected during these studies. This would involve obtaining sufficient PK measurements to ensure that the parent compound reaches the target and drives the pharmacologic read outs—in particular in models or paradigms necessitating chronic administration. Conventional chronic toxicity testing would not be necessary, but the efficacy experiments should be associated with a behavioral assessment (Irwin-like) of adverse effects. Detailed preclinical good laboratory practice (GLP) testing of drug metabolism and PK, safety, and toxicology required to generate the data package necessary to begin first in man/phase I clinical studies, including chronic GLP toxicity testing in rodent and nonrodent species, could be undertaken before, after or in parallel with the target validation phase II multicenter preclinical studies, and are not an integral component of this model.

In contrast, phase II studies of nonpharmacologic therapies may include tolerability and toxicity data that are not obtained at other phases of validation and development.

Selection of laboratories to be involved in preclinical studies

It is proposed that selection of laboratories is undertaken by a coordinating center from a database of laboratories that have volunteered to be involved in phase II preclinical studies and have been accredited based on criteria including:

  1. Track record of high quality preclinical studies of therapeutics for epilepsy, evidenced by peer-reviewed publications, site inspection, and record of previous participation in phase II studies.
  2. Availability and experience in the animal models selected for the study.
  3. Capacity to undertake the required number of studies within current workload (estimated 10–20 animals in 1–3 models).
  4. Capacity to store and account for investigational therapies to be tested.
  5. No significant conflicts of interest.

The lead investigators at participating laboratories testing a significant number of subjects would be offered authorship on publications of the trial, as is currently the case with clinical trials.


It is recognized that these phase II multicenter preclinical studies will be more expensive and resource and time intensive than traditional preclinical studies. However, the cost should be significantly less than that of phase II/III clinical studies, and therefore overall should prove to be cost-effective to the sponsor.

The funding model will likely require a combination of government and industry funding. The government funding would be best utilized to establish the basic structures, protocols, laboratory credentialing, databases, and so on. Industry or venture capital (i.e., the sponsor) would fund the primary costs of undertaking the study, potentially supplemented by grants from government and philanthropy. Participating investigators would receive “investigator payments” to cover the costs of undertaking the studies in the laboratories, and appropriate infrastructure costs.

The intellectual property rights for the therapies would remain owned by the “sponsor” of the compound.

Validation of the “phase II” multicenter preclinical drug trial model

It is important to acknowledge that it is currently an unproven hypothesis that a positive result from phase II multicenter preclinical studies, as proposed herein, will result in an improved success of translation into positive clinical studies. A validation study should be performed to prove the predictive value of this model. This could be done by testing drugs that have been successful in clinical trials and practice (e.g., levetiracetam) in a phase II multicenter preclinical study. Ideally, in addition to comparing the drug of interest with vehicle control, the studies should compare with a drug that was promising in traditional “phase I” preclinical studies but showed limited efficacy in subsequent clinical trials (e.g., carisbamate) or in clinical practice (e.g., gabapentin).

Criteria to Move from “Phase II” Preclinical into Clinical Development

To be recommended to proceed from phase II preclinical studies into clinical studies, therapies should ideally show better efficacy than existing therapies on drug-resistant seizures, or efficacy in clinical practice for end points representing current “gaps in epilepsy care” for which no effective therapy exists (i.e., epileptogenesis, disease modification, and comorbidities). It would also be expected that the new therapy had favorable toxicity and PK profiles. Altogether, it is therefore proposed that preclinical studies may include:

  1. Phase I preclinical: Proof-of-concept and early preclinical studies (mainly testing efficacy, with some information on drug metabolism, PK, and minimal toxicity battery (Galanopoulou et al., 2013, in this Supplement).
  2. Phase II preclinical: As proposed here for multicenter studies.
  3. IND-enabling studies: Main focus on safety-toxicity and on providing information for the design of first-in-human studies.

It should be noted again that we do not advocate that all therapies must be tested in phase II preclinical trials prior to proceeding to clinical trials, as there may be circumstances where the therapy sponsors believe there is already enough preclinical evidence for efficacy based on traditional “phase I” preclinical studies.


In this article we have proposed and outlined a model for “phase II” multicenter preclinical trials with the aim of improving the preclinical evidence base for the efficacy of new epilepsy therapies, in particular to address the major current gaps in epilepsy care: drug-resistant epilepsy, disease modification, and comorbidities. The primary goals are to “derisk” clinical trials and to motivate industry and other funders to invest in the development of new therapies for people with epilepsy. The successful implementation of this model would require an alignment and collaboration from government agencies, industry, and academia.


This article has derived from the work of “Subgroup 7” of the London Workshop of the ILAE/AES Working Groups joint meeting to optimize preclinical epilepsy therapy discovery held 28–29th September 2012 in the Custom House Hotel, London, UK. The support for this Workshop of the International League Against Epilepsy (ILAE), American Epilepsy Society (AES), Citizens United for Research in Epilepsy (CURE), Epilepsy Therapy Project, and Autism Speaks for the London workshop on optimization of preclinical epilepsy therapy discovery is gratefully acknowledged by all authors.


Terence J. O'Brien has received research grant support from the NHMRC (Australia), the Royal Melbourne Hospital Neuroscience Foundation, Sanofi-Aventis, UCB, SciGen, GlaxoSmithKline, Novartis, and Janssen-Cilag; and speaker honorarium from Sanofi-Aventis, UCB, SciGen, GlaxoSmithKline, and Janssen-Cilag. Henrik Klitgaard is an employee at UCB Pharma. HL has received funding from the German Federal Ministry of Education and Research (BMBF), German Research Foundation (DFG), European Commission, UCB, Sanofi-Aventis, travel support or honoraria for speaker's engagement and consulting from Cyberonics, Desitin, Eisai, GlaxoSmithKline, Medtronic, Pfizer, UCB, and Valeant. Michele Simonato has received research funding from the European Union (Epixchange), the Italian Ministry for Research and the University (Prin 2010-11), GlaxoSmithKline, Chiesi Pharmaceuticals (Italy), Sanofi-Synthelabo, and Schering-Plough. Matthew C. Walker, has received research funding support from Medical Research Council (UK), Wellcome Trust, NC3Rs, Epilepsy Research UK, National Institute of Health Research (UK), UCLH and UCL biomedical research centre and European Union and speaker honorarium from UCB, GlaxoSmithKline, Viropharma and Eisai. Elinor Ben-Menachem has been a Consultant for Eisai, BIal, Lundbeck, Electrocore, Janssen Cilag, and Biocontrol. She receives research support from UCB, Bial, and Eisai Inc. and Västra Götaland Region. She is Chief Editor of Acta Neurologica Scandinavica. Edward H. Bertram III has received research funding from the National Institutes of Health.

We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.