final report, to make the following quite remark-
able statement:
‘No drug which is pharmacologically effective is
entirely without hazard. The hazard may be insig-
nificant or may be acceptable in relation to the
drug’s therapeutic action. Furthermore, not all
hazards can be known before a drug is marketed:
neither tests in animals nor clinical trials in
patients will always reveal all the possible side
effects of a drug. These may only be known when
the drug has seen administered to large numbers of
patients over considerable periods of time’.
Thus, we can try to prevent these drug disasters
to the maximum extent possible, we can try to
identify and diagnose them at the earliest possible
date (and so minimize the number of patients hurt
or disadvantaged), but, in the present state of scien-
tific knowledge, we cannot eliminate drug with-
drawals due to unexpected toxicity.
A list of 39 drugs withdrawn due to major safety
concerns in the United Kingdom between 1975 and
2005 has recently been published elsewhere (Mann,
2005). This list can be summarized in Table 44.1.
It is notable that 11 (28%) of these withdrawals
were due to causes which appeared only once in the
30 years considered in the table. Examples are
polidexide, withdrawn in 1975 as the formulation
contained impurities, and coproxamol, withdrawn
in 2005 as it was frequently used in suicides. The
most common single cause of withdrawal was hepa-
totoxicity (nine drugs withdrawn), but adverse car-
diovascular events, if grouped with arrhythmias,
were equally conspicuous. Drugs which challenge
liver function would seem poor candidates for
development and it seems reasonable to suggest
that prolongation of the QTc interval clearly needs
to be carefully excluded before new drugs move
very far along their path of development.44.2 Prevention is better
than cureIt is now well understood that post-marketing sur-
veillance (PMS) must be undertaken when newly
licensed drugs enter everyday clinical usage. The
data available at the time of marketing speak far
more fully to the efficacy and quality of the drug
than they speak to its safety. This is because the
number of patients included in the clinical trials is
small compared with the number of patients who
can be exposed once the drug is available for pre-
scription; additionally, the populations are very
different, because the prelaunch program will
have been conducted in patients with only one
disease, whereas the drug, once licensed, will
often be used in an older population of patients
with more than one disease. Thus, the PMS pro-
gram should emphasize populations covering the
age ranges appropriate to the indications for the
drug; it should also include appropriate drug–drug
interaction studies and studies in any special popu-
lations, such as children, who may receive the drug
once it is in everyday clinical usage. Clearly, the
PMS program needs to include studies intended to
resolve any queries or questions that have been
noted in the animal safety evaluation studies or
the earlier studies in man.
A great deal of guidance is now available and
is, in its essentials, common to all areas of theTable 44.1 Reasons for 39 Drug Withdrawals: UK
1975–2005
Reason for Number of
withdrawal drugs withdrawn
Hepatotoxicity^19
Cardiovascular^25
Arrhythmias^34
Skin and mucus membrane lesions^44
Gastrointestinal lesions^52
Blood disorders^62
Anaphylaxis^72
Others (singles only)^811
Total 39
Key:
1: benoxaprofen, clomacran phosphate, perhexiline, dilevalol,
pemolin, troglitazone, tolcapone, trovafloxacin, kava-kava.
2: fenfluramine, dexfenfluramine, droperidol, amfepramone
phenteramine, rofecoxib.
3: terodiline, sertindole (license restored), grepafloxacin,
cisapride.
4: practolol, fenclofenac, feprazone, valdecoxib.
5: indoprofen, Osmosin.
6: nomifensine, remoxipride.
7: zomepirac, Althesin.
8: polidexide, zimeldine, suprofen, metiprannolol eye drops,
triazolam, temafloxacin, remoxipride, mibefradil, Alec, ceri-
vastatin, coproxamol.
580 CH44 DRUG WITHDRAWALS FROM THE MARKET – CAUSES AND CONSEQUENCES