Risk-benefit evaluation in approval of traditional and innovative drugs

Oral presentation

Over the last few years, the natural history of diseases has changed little. While traditional drugs had traditional problems, novel medicines have often brought with them additional novel problems. Good examples of these are the nephrotoxic and alleged pro-ischaemic effects of COX-2 selective inhibitors, Churg-Strauss syndrome in association with leukotriene receptor cysLT1 antagonists and unusual CNS events and infections associated with some TNF-a blockers.

The risk/benefit of drugs active at the well known pharmacological targets compared to those acting at novel targets are often not better appreciated. While ACE inhibitors and some (-blockers have greatly reduced morbidity and mortality in heart failure, similar benefits have yet to be observed with endothelin receptors antagonists. The evaluation of risk/benefit of any medicine, novel or traditional, requires a number of factors to be considered.

On the benefit side, it is important that the drug is shown to have beneficial effects in terms of hard clinical endpoints such as morbidity, survival and quality of life.

As far as the risk is concerned, one needs to consider the frequency, severity and reversibility of the risk. It is also important to consider how easy it is in clinical practice to detect or diagnose the risk. The overall risk may be a composite of a number of rare events. This is well illustrated by perhexiline associated with neuropathy, hepatotoxicity and severe weight loss, temafloxacin associate with severe hypoglycaemia, haemolytic anaemia, other blood cell abnormalities, nephrotoxicity requiring renal dialysis, hepatic dysfunction and severe allergic reactions with respiratory distress.

About 75% of adverse drug reactions are type A pharmacological adverse events that are concentration or exposure related. A more striking example of the dose dependency of risk/benefit evaluation is evident for drugs whose pharmacokinetics or pharmacodynamics are influenced by genetic polymorphisms. Therefore, more critical is the appreciation that risk is not fixed but varies with dose. Traditionally, drugs have been investigated and recommended for clinical use at higher than optimal doses. This is most vividly illustrated in the recent times with cerivastatin. Originally approved at 0.3mg daily dose, the maximum recommended daily dose at the time of its withdrawal was 0.8mg. It must be questioned if the benefits had increased proportionally.

Risk/benefit may also be influenced by a number of other factors. Apart from the dose, these include the age, ethnicity or genotype of the patient, co-medications and co-morbidity.

The perception of risk benefit also varies with local expectations and medical practice. The availability of alternatives also greatly influence the extent to which one is prepared to accept risks associated with a particular drug. There may be differential emphasis on efficacy (eg in the US) or safety (eg in Japan). Between January 1995 and July 1999, 135 full applications were determined by CPMP. Of these, 38 had a negative outcome.

By January 2000, FDA had authorised 13 (34%) of these 38 applications. Following the reports of torsade de pointes, while levacetylmethadol was suspended in the EU, the drug was relegated to second line treatment in the US.

Even within the European Union, these differences become obvious when one considers the withdrawal of products going through the mutual recognition procedure.

Have something to say? Post a comment on this article!