The Wall Street Journal reports (Jan 2, 2009) that a new collaboration between pharmaceuticals giant Pfizer and two Boston hospitals will test whether computerized patient records can boost reporting of adverse drug reactions (ADRs) making it a routine part of filling out electronic patient charts.
Some time ago (Catalyst column, ChemWeb.com, June 1998), I discussed the implications of the more than 100,000 deaths in the US each year allegedly caused by patients’ reactions to their medication – three times the number killed in car accidents. So-called adverse drug reactions (ADRs) are, estimated to be the fourth biggest killer in the US after heart disease, cancer and stroke. Recently, there has been an upsurge of interest in ADRs and calls in the US for an independent body to be established to make control of drugs once they have passed though the regulatory process easier and save lives.
That 100,000 is just a statistic of course, except for those patients and their loved ones affected. Every drug has side-effects and although they do not exist through malicious design, one can perhaps see that the drug R&D process is not perfect.
A pharmaceutical company for reasons of economics and politics cannot possibly study the effects of every putative drug on every ‘type’ of individual in the different circumstances in which it might be used. This is where medication monitoring services come in handy. Pharmacogenomics and personalised medicine that focus on each patient’s single nucleotide polymorphisms (SNPs) may remedy this. But, despite the emergence of inexpensive genomics and predictions of the $1000 genome, this is still true when it comes to administering to the elderly and children as they can be more sensitive than the proverbial adult. Moreover, in the supposedly clinically correct environment of the hospital there are likely to be even more exacerbating factors at work for each individual patient than there might be for a patient with a straightforward bacterial infection, say.
An individual’s genome may be at the root of a particular type of adverse drug reaction. As Catalyst discussed early in 1998. Ten percent of Caucasians and about two percent of Chinese people cannot metabolise the analgesic (painkiller) codeine into its active form, morphine. The drug therefore simply does not ‘work’ for them. The problem boils down to those patients lacking the gene for the liver enzyme CYP2D6 responsible for the conversion. This particular effect was discovered by Alastair Wood a clinical pharmacologist at Vanderbilt University in Nashville, Tennessee. The drug having no apparent effect might lead the GP to prescribe a higher, perhaps intolerable dose. For a Chinese person lacking CYP2D6 the result can be severe nausea.
CYP2D6 metabolises a variety of drugs in addition to codeine, for instance, the antihypertensive propranolol (Inderal), propafenone (Rythmol), for heart arrhythmia, and many of the tricyclic antidepressants. In these cases though people lacking CYP2D6 actually experience an exaggerated effect as the active form stays in their system longer.
In the hospital environment, muscle relaxants used in anaesthesia can be a particular problem for some patients, because they have a faulty gene for the enzyme, butyrylcholinesterase, that would naturally metabolise that drug. For example, succinylcholine stops patients breathing during surgery, this is fine while mechanical ventilation is continued but for some patients the apnoea does not cease and they can die. Peculiar peak concentrations of the TB drug isoniazid have been seen with some patients and have been correlated with a faulty N-acetyltransferase.
In fact, there are many, many variations in drug response that have been recognised and the pharmaceutical companies are becoming well aware of the potential for profit these variations might bring if they can develop drugs tailored to an individual’s genome. The National Institutes of Health in the US has also recognised the potential for improving medicine and is in the process of establishing a Pharmacogenetic Polymorphic Variants Resource database for genes encoding proteins that determine variations in drug responses.
Pharmacogenomics ties in closely with the reporting of adverse drug reactions, although not all ADRs are due to genes. The anti-obesity drugs dexfenfluramine and fenfluramine which are often taken in combination with phentermine – as fen/phen – caused serious ADRs in the form of major heart valve problems in 31% of patients taking the combined medication. The eventual withdrawal of the drug once the problem was widely recognised and publicly known was swift but fenfluramine had been on the market 24 years.
However, while the voluntary reporting of ADRs is fairly common within the medical profession their existence is not well known. Indeed, aside from mentioning a few cursory side effects doctors are often unaware of potentially serious reactions to particular drugs and this is compounded by the fact that all this reporting of ADRs is purely voluntary with the onus on the pharmaceutical companies. As such, there are many people unfairly affected by these drugs and there are actions against pharmaceutical companies like the lawsuit against Levaquin, Actos, etc. It took twelve years before the antihistamine drug used by countless hayfever sufferers every summer was withdrawn in preference to its safer metabolite. The major ADR of terfenadine is potentially fatal heart arrhythmia especially in users taking certain antibiotics at the same time.
A group of medical scientists led by Alastair Wood, published a paper in the New England Journal of Medicine (1998, 339, 1851) calling for an independent drug safety board to be established to keep tabs on ADRs. This body would be there to help protect patients as well as ensuring that medical practitioners were made fully aware of the putative hazards of the countless drugs they prescribe.
According to Wood and his colleagues, ADRs are a serious cause of patient morbidity and mortality. They make the point that there have been independent bodies in place to investigate the likes of plane crashes, train and major traffic incidents, chemical and radiation accidents for many years. These bodies can make recommendations to prevent similar serious episodes happening again following an accident. But, there is no organisation with responsibility for monitoring ADRs and to ensure proposals put forward following an investigation are taken on board.
The ad hoc approach to reporting of ADRs and reactions to drug products seems at odds with the fact that we have Internet and information technology available. Wood and his colleagues say that for all this technology it is remarkable that little use is made of it for drug surveillance to help avoid the huge numbers of deaths that occur. The likes of terfenadine and phen-fen which do end up being withdrawn by the FDA are few and far between and the evidence on which the decision is based while strong is not often in the form of formal statistical analysis. One of the problems is that the US Food and Drug Administration (FDA) does not have the resources to carry this out nor is it in the interests of the pharmaceutical marketers to gather such data.
Wood and his colleagues believe that the solution to the problem is to make this surveillance obligatory through the creation of a body independent of the agency that carries out drug approvals – the FDA. A second, independent body would help avoid conflicts of interest, in that the FDA would not have to investigate problems with drugs it had approved! In their paper in the NEJM the authors state,
We must expect that predicted and unpredicted adverse events from drugs will continue to occur. If we accept that the true safety profile of a new drug is dependent on the experiment that necessarily follows the drug’s release into the marketplace, then we must fund and implement mechanisms to ensure that the experiment is properly monitored, the data appropriately analysed, and the conclusions disseminated rapidly.
Clinical trials can involve a few thousand people, once approved, millions may take it soon after especially now that TV marketing is available in the US to the companies.
Not all ADRs are lethal, just adverse, and some are simply unavoidable because of the individual circumstances in which a drug is administered. They may be unpredictable and unavoidable in some cases but once an ADR occurs the medical community should be made aware of the risks as soon as possible so that better judgements about prescribing a drug can be made and ADRs pushed right down that list of causes of death.
This original version of this article appeared in my Catalyst column in ChemWeb’s The Alchemist in March 1999 before Vioxx, pre TGN1412, and only the intro has been updated January 2009.
