Clinical pharmacology and drug regulation

The role of clinical pharmacology in drug regulation may seem blindingly obvious—for clinical pharmacology explores how drugs affect man, including the systems on which they act as well as how man affects drugs—absorbing, distributing and eliminating them.

In order to protect the public health, the drug regulator must show that products recommended for licensing are safe and effective and of such a quality that their safety and efficacy are not compromised. The regulator must also ensure that marketed medical products are promoted accurately and objectively by manufacturers to both health care professionals and to the public.

I wish to discuss three aspects of drug regulation which have direct implications for the clinical pharmacologist. Firstly, measurement of efficacy, secondly, assessment of safety and thirdly, the concept of risk and benefit.

Efficacy

Efficacy has historically not been the main preoccupation of drug regulatory authorities in this country or elsewhere. Sir Derrick Dunlop, the first Chairman of the (then) UK Committee on Safety of Drugs wrote in 1967 ‘The Committee’s remit does not impose upon it any responsibility to consider the efficacy of drugs except insofar as their safety is concerned’. Perhaps, in the wake of recent problems with thalidomide, this was not altogether surprising. It was not until the Medicines Act was introduced in 1972 that evidence of efficacy as well as of safety was required as a condition for granting a product licence. Interestingly, as Jefferys and colleagues discuss [1] over the past 20 years only one drug has been withdrawn from the UK market because of lack of efficacy. Oxalinic acid, a urinary antiseptic was shown in 1979 to cause extremely rapid resistance in hitherto sensitive organisms and its licence was suspended. This has to be compared with the twenty-two products whose licences were withdrawn because of safety concerns over a similar period.

The bedrock measurement of drug efficacy, which many clinical pharmacologists have helped plan and participate in is the randomised controlled clinical trial and this continues to serve the public health well. Where controversy does arise, however, is in the end points of efficacy which the clinical trialist may choose to adopt in his trial. Where mortality or appropriate clinical endpoints e.g. stroke or myocardial infarction exist, little argument occurs between interested parties. However, the use of such measurements of efficacy usually implies long trials with large numbers of patients. This has thus led to a debate between industry and the drug regulator (with the clinial pharmacologist as an interested observer and participant) of the value of what has become known as surrogate markers of efficacy. A surrogate marker is defined as an accurate indicator of disease progression that does not rely on clinical findings. Such markers are used with varying degrees of confidence throughout clinical medicine and in drug development. For example, blood pressure is widely accepted as a surrogate for stroke and myocardial infarction. Tumour enlargement is used as a surrogate for the effect of anticancer treatment, even though it may not predict survival. Intraocular pressure measurement is used as a clear surrogate for the clinical consequences of glaucoma. Bone density is used as a surrogate for the development of osteoporotic fractures, and for the assessment of drugs which influence osteoporosis.

For a surrogate to be used as the sole end point of measurement of efficacy for regulatory purposes, it must exhibit a high degree of robustness. In each of the instances cited above, a good correlation has been shown to exist between the surrogate and an accepted clinical end point. Regulatory approval of a drug based on an inappropriate surrogate may lead to many patients receiving ineffective or even harmful therapy.

A therapeutic area where this debate has been most pointed is in the treatment of HIV disease. The UK licensing authority has been more reluctant than its American or European counterparts to grant licences for anti HIV drugs unless the agent had shown significant changes in clinical end points such as difference in survival, progression to an AIDS defining clinical event or reduction in the incidence of opportunistic infection. Industry wishes to speed up the licence approval process by submitting surrogate data such as changes in CD₄ count, or in viral load, i.e. markers of the state of the immune system or the amount of virus present as evidence of anti HIV efficacy. In this context it is worthwhile remembering that the ideal surrogate marker should a) be biologically plausible, b) be detectable in most subjects at all stages of the disease, c) show changes towards normal when an effective agent is given, d) predict the ultimate clinical response in patients taking the drug, e) discriminate between patients who will do well and those who will do badly. In HIV disease, intense debate continues as to the value of the various surrogate markers and which conform to these five criteria. As ever, drug regulation must follow science, and science has not yet made up its mind on this issue.

Safety

Adverse reactions to drugs range from effects on a single physiological control system to multiple organ failure, and their severity can vary from the apparently trivial to the rapidly fatal. Clinical pharmacology is concerned with risk assessment in drug therapy and as the science of toxicology develops, new classifications of adverse reactions emerge. As with considerations of efficacy, regulation of drug safety must follow this scientific progress. Adverse reactions which cause most problems to the clinician and the regulator are the so called Type B (bizarre effects due to drug action other than those of a therapeutic nature) rather than the common Type A (augmented response predictable from the drug’s pharmacology). Clinical and biochemical pharmacology has shown that these infrequent but extremely dangerous Type B reactions may be due to either parent drug or to a toxicologically reactive metabolite. Where an imbalance occurs between the formation of such a metabolite and its detoxification, toxicity may result. If the reactive metabolite binds covalently to essential macromolecules, or to DNA, or acts as a hapten, tissue damage may occur and clinical toxicity ensues. Factors influencing individual susceptibility to these idosyncratic forms of drug toxicity are poorly understood, but probably include genetic, metabolic and immunological factors. With respect to those idiosyncratic reactions mediated via a hapten causing immune mediated toxicity, it may even be difficult to distinguish the effect of the immune response from that of the direct toxicity of the reactive metabolite. Drugs of all classes may cause such reactions and they pose great problems for the pharmacologist and the regulator. Among the twenty two drugs whose licences were withdrawn for safety reasons over the past twenty years, idiosyncratic toxicity such as anaphylaxis, hepatotoxicity and haemolytic anaemia are frequently the basis for the regulatory decision. Only by understanding the biochemical and chemical basis of these effects can drug design be improved to minimise drug toxicity.

Safety evaluation in drug regulation depends on three strategies—the control of quality, rigorous premarketing safety studies and post marketing surveillance. Quality is controlled in relation to both manufacture and wholesale selling and only one licensed drug (penbutolol) has been withdrawn from the UK market over the past twenty years because of quality problems. Premarketing safety evaluation is based on evidence from both preclinical and clinical studies. At best preclinical studies may predict the majority of so called Type A (augmented) adverse reactions, but most attempts to correlate findings during human use with those during preclinical toxicity have met with only limited success. Premarketing safety evaluation in man is of the highest importance at all stages of clinical drug development, but because of the relatively small numbers of patients studied (some 1000–1500) its value for identifying the rare but life threatening Type B idiosyncratic reactions described above is obviously limited. Post marketing surveillance is thus the key stratagem for such identification, and vigorous debate is still enjoined between the drug regulator and the clinical pharmacologist as to the relative value and contribution of the various surveillance techniques available.

A frequently forgotten aspect of clinical pharmacology and drug regulation, which in some respects is the most important of all, is examination of the Summary of Product Characteristics of a new drug. This document provides essential information for the prescriber and is the basis for patient instructions and prescribing guidelines. Not only must this document be accurate but it also needs to be easily understood.

Risk and benefit

The concept of balancing risk and benefit is not one which the public easily entertains. Too often, the same standards of safety are expected for an agent which may be life saving after a myocardial infarction and one which relieves itch. Drugs are too often still envisaged as silver bullets. Both the clinical pharmacologist and the drug regulator know this not to be so, but perceptions on the appropriate balance of risk and benefit vary widely, including nationally. While formal analysis of risk and benefit for a particular drug can be carried out (e.g. measurement of lives saved against morbidity); comparative assessment is probably slightly more useful (i.e. comparing risk and benefit with similar drugs in therapeutic area), but the regulator will eventually come to rely on his own innate judgement to make his assessment. Confronted by evidence of efficacy of whose basis he may even be uncertain, together with limited preclinical and clinical safety data, and no postmarketing experience on which to assess risk, licensing decisions are often made on as much a judgmental as a scientific basis.