As Neil Parish, a British member of the European Parliament and one of the cochairs of the meeting, said: "I get more letters and emails from people concerned about animal welfare than anything else I deal with."

Among the broad areas of agreement, the participants at the meeting concurred on the key principle that less animal testing must not impact negatively on patient safety. There was also agreement that trial subjects must be informed about all relevant tests performed in animals and about what the results indicate about risks potentially connected to the administration of the new drug. In addition, any limitations of the tests in terms of their predictability of safety need to be pointed out.

There was a general feeling that more harmonization of requirements both within the EU and among the EU, the United States, and Japan would help to reduce the number of animals used in experiments. Ethics committees considering first-in-man trials should have access to acknowledged first-in-man experts, it was agreed—in view of the obvious fact that a reduction in animal testing would put additional responsibility on ethics committees.

But there were plenty of divergences too. The meeting indicated that some find the current regulatory framework appropriate, pointing to the low incidence of serious adverse drug reactions in trials. Others, however, feel that the regulatory requirements for doing tests in animals are not always justified, and urge continuous review of the current toolbox of experiments in line with progress in biomedical research—and, if appropriate, reduced rather than enlarged.

But while it would clearly be acceptable to reduce animal use in cases where no suitable animal models exist, it is not at all clear how much animal experimentation is redundant now. This makes it difficult for trial sponsors, who might wish to stop doing studies in animals that are formally required but known to be useless in a particular situation. Since they have to justify whenever they diverge from the toolbox, regulators should offer more extensive guidance in early discussion with trial sponsors about animal studies, particularly through the scientific advice procedure that the European Medicines Agency provides.

Child's play

New regulations do work—sometimes. At the end of September, the European Medicines Agency made its first recommendation for the use of a centrally authorized medicine in children on the basis of trial data generated in accordance with an agreed pediatric investigation plan (PIP). This is the first example of the changes brought in by the European Union's recent pediatric regulation, designed to benefit children in Europe by promoting the availability of medicines specifically tested for their use.

PIPs, one of the tools created when the new rule entered into force last year, are drug development plans for ensuring that the necessary data are obtained through studies in children, when it is safe to do so, to support the authorization of the medicine for children. These plans must be agreed in advance by the Agency's pediatric committee, and are legally binding for companies developing products for use in the EU.

Using the data generated as part of a PIP, Merck Sharp & Dohme applied for an extension of indication for use in children for its Cancidas (caspofungin), already authorized for the treatment of adults with severe fungal infections. The Medicines Agency adopted an initial decision on the PIP for Cancidas in February 2008 and a modification in May. The Agency's Committee for Medicinal Products for Human Use concluded in September that the data on the quality, safety, and efficacy presented support this extension.

Waiving not drowning

By coincidence, it was in September too that the EU published its formal guidance on the content of pediatric investigation plan proposals, and on how to seek waivers from the obligation to present a plan. This acknowledges that the "one-size-fits-all" guidance is bound to lead to some gaps in proposals, since the amount of information available "will differ substantially depending on whether a medicinal product is in early clinical development or already has a marketing authorization and is being investigated for new or extended uses."

However, it does insist that where relevant information is available, it should be included in the application, "whether favorable or unfavorable to the product." This includes details of any incomplete or discontinued pharmacotoxicological test or clinical study or trial, and any completed trials concerning indications not covered by the application. Applicants are urged to consider whether there may be a therapeutic need for the medicinal product in each pediatric subset when they are drafting proposals.

Just as important as PIPs is the subject of waivers from the obligation to present a PIP. Here the new guidance lays down that a waiver may be issued for one or more specified subsets of the pediatric population, or to one or more specified therapeutic indications, or to a combination of both, but only if there is a clear definition of scope in terms of pediatric subset and indication—and of route of administration and pharmaceutical form.

There are very few free lunches in the guidance. Where the therapeutic indication and the subset of the pediatric population are already covered by a class waiver, no product-specific waiver is required, it concedes. But where only a partial class waiver exists, the product-specific waiver should clearly refer to the class waiver when specifying the scope of the product-specific waiver. In addition, companies are "encouraged" to inform the Agency's Pediatric Committee when new information becomes available, which suggests that a class or product specific waiver should be reviewed. Under such a welter of requirements, it will be no surprise to find many sponsors and investigators who feel they are drowning rather than waiving.

Peter O'Donnell is a freelance journalist who specializes in European health affairs and is based in Brussels, Belgium.