Knocking at the door of harmonisation

The US and EU have made progress towards harmonising their new drug authorisation systems, but mutual recognition remains elusive, writes Richard Kingham, a partner at international law firm Covington & Burling

The standards for approval of medicines are now essentially the same on both sides of the Atlantic, as are the requirements for non-clinical tests and clinical trials to meet those standards. Marketing authorisation applications for most products will soon be submitted in a standardised format in the US and the EU. Nevertheless, the philosophies that underlie the medicines review process remain strikingly different.

Until the mid-1970s, the US Food and Drug Administration functioned as if the US existed in a vacuum. Reports of clinical trials conducted outside the US could not be submitted to fulfil even the most basic requirements of the US marketing authorisation process. Even when a product had been used extensively in medical practice in Europe, the FDA was less concerned about foreign clinical experience than the results of clinical trials conducted in the US.

An early breakthrough related to non-clinical studies. In the 1970s, responding to a series of high-profile cases of fraud by animal testing laboratories, the FDA instituted elaborate rules for good laboratory practice (GLP). Shortly afterwards, the OECD issued broadly similar requirements, which were implemented in Europe, and the FDA entered into memoranda of understanding with several European countries on mutual recognition of GLP inspections.

In the mid-1970s, the FDA also liberalised its rules on acceptance of foreign clinical data for early phases of clinical research, but the agency continued to reject pivotal studies that were not carried out under US investigational new drug applications (INDs) until the mid-1980s. Even today, there is a greater chance that the FDA will demand site inspections for pivotal trials carried out in Europe rather than the US.

It is possible that EU GCP inspections, which will be required when the clinical trials directive takes effect in 2004, will increase FDA's confidence in reports of clinical trials from Europe. But for the time being prudent pharmaceutical companies will aim to carry out at least some of their pivotal clinical trials in the US.

The US pharmaceutical industry has often argued that the FDA has much to learn from Europe. In the early 1980s, it contended that review times in the most efficient EU countries were far shorter than in the US, with no apparent loss in protection of public health, while procedures for approval of clinical trials were significantly less bureaucratic.

Although the FDA resisted making fundamental changes to its review procedures, the European experience helped put political pressure on the agency to streamline its approach to the new drug application (NDA) and investigational new drug processes.

The pace of harmonisation was greatly accelerated in the late 1980s and 1990s through the efforts of the International Conference on Harmonisation (ICH), which brought together representatives of the medicines authorities in the EU, US and Japan, as well as participants from other jurisdictions and from industry. Working parties on safety, efficacy, quality and other topics reached agreement on a wide variety of guidelines dealing with nearly every aspect of the medicines approval process. Most important, the governments of the participating countries were committed to amending their national policies and regulations to implement the standards agreed through the ICH process.

Recently, the ICH agreed a standardised format for marketing authorisation applications (the 'common technical document', or CTD). Although parts of the CTD must be tailored to local requirements, the bulk of a marketing authorisation application can now be prepared for submission in essentially the same form to the authorities in the three largest pharmaceutical markets. The CTD will soon be mandatory in Europe. Although its use remains voluntary in the US, the FDA is strongly encouraging most applicants to make filings in the ICH format.

Despite these successes in the effort to achieve harmonisation, mutual recognition of marketing authorisations between the US and the EU has proved elusive. This should not be surprising.

Even within the EU, it took many years for member states to accept one another's approval decisions.

The current mutual recognition system, established in 1995, ultimately depends on the power to impose binding, Community-wide decisions. National authorities still disagree on approval decisions, and applicants are often forced to withdraw submissions in one or more member states to avoid lengthy and unpredictable arbitration proceedings.

fundamental differences

The differences between the US and Europe are, if anything, more fundamental. They may be traced in part to differing histories of drug regulation in the US and Europe, as well as differences in the political systems within which medicines agencies must operate.

The driving force in European drug regulation for the past decade has not actually been consumer protection; it has been the economic goal of perfecting the internal market of the European Union. The institutions created to achieve that objective (the European Agency for the Evaluation of Medicinal Products, or EMEA, and the Committee for Proprietary Medicinal Products, or CPMP) concentrate mainly on scientific and medical issues, leaving enforcement to the national authorities.

In comparison, the FDA operates in a fishbowl. The Freedom of Information Act requires nearly every document created by the FDA and submitted to it, to be disclosed to any interested member of the public. Most meetings of FDA advisory committees are required to be held in public, and committee sessions are regularly attended by news reporters and financial analysts. US laws protect whistle-blowers, who regularly leak sensitive documents, and public interest advocates are more active and more strident, than in many European countries. The US system of divided government ensures that Congress acts as an independent watchdog on the FDA, and judicial challenges to FDA actions are more frequent than in most European countries.

inherent scepticism

These factors have made the FDA inherently conservative and sceptical towards the industries it regulates, with significant implications for the medicines review process. Because FDA officials do not trust industry, they insist on re-tabulating and re-analysing the data in NDAs. Often, they demand not only all data collected in clinical trials but also the case report forms from which those data are drawn. When they spot an anomaly, they are quicker than their European counterparts to call for inspections of records at the site of the clinical investigation. In effect, the EU reviews a marketing authorisation application from the top down, while FDA reviews it from the bottom up.

Throughout the 1970s and 1980s, all efforts by the pharmaceutical industry to convince the FDA to accept 'certified summaries' of tests, or expert reports of the type that accompany marketing authorisation applications in the EU, met with failure. By the early 1990s, the US industry concluded that it would never convince the FDA to adopt the European approach to the review process, and decided instead to give the agency the resources to carry out reviews using its traditional approach. The Prescription Drug User Fee Act of 1992 imposed registration fees that gave the FDA money to hire hundreds of new medical reviewers, in exchange for shortened review times. That law, renewed in 1997 and due to be re-enacted this year, has probably had more real impact on US review times than any other measure in the last 30 years. For much the same reason, US manufacturers were quick to embrace electronic marketing authorisation submissions, since they facilitate the re-tabulations and re-analyses that FDA reviewers prefer.

A second major difference between the US and EU reviewing processes results from administrative organisation. From the earliest stages of clinical research through the approval process and the post-marketing stage, the FDA assigns a drug to a reviewing division on the basis of the product's therapeutic use. Although a reviewing division is supported by other units, it retains primary responsibility for questions relating to the safety, efficacy and labelling of a drug product throughout its regulatory lifespan.

In contrast, most medicines authorities in the EU separate these functions. Clinical trials are approved by a different group from the one that reviews marketing authorisation applications, and yet another organisation is responsible for post-marketing surveillance.

diffusion of responsibility

There is even greater diffusion of responsibility when medicines are reviewed at the Community level. Clinical trials are regulated separately by national authorities in each member state where studies are conducted. Although companies can apply to the CPMP for advice on clinical trial design and other issues, the process is bureaucratic, and there is no assurance that the persons giving the advice will play a key role when the marketing authorisation application is reviewed.

The Community review process is divided between two rapporteurs, who assign tasks to reviewers in two national agencies. The result is a consensus decision in which no regulator has a clear professional stake. If safety issues arise after a product enters the market, a company can find itself very much alone, having few, if any, advocates in the CPMP.

These fundamental differences in approach to the review process, coupled with senior FDA officials' political accountability for approval decisions, make it highly unlikely that the US will accept any system of international mutual recognition of marketing authorisations.

Experience with the US-EU mutual recognition agreement on pharmaceutical good manufacturing practice (GMP) inspections suggests, moreover, that the FDA would probably resist any system that was imposed by politicians.

Acting under pressure from the Clinton Administration, the agency agreed in principle to the GMP agreement, but insisted on a lengthy 'confidence-building' period, with the right to limit the number of EU member states whose systems were deemed equivalent to the FDA's. It remains to be seen whether the agreement will work in practice, but experience to date suggests that the results will be less substantial than industry hoped.

The FDA simply cannot trust another medicines agency to review applications, and make decisions as to benefits and risks and labelling, in a manner that it can defend to the US Congress and the American public. For the time being, therefore, it is likely that final decisions will be made separately in the US and Europe.