56●Pharmacokinetic changes 56●Pharmacodynamic changes 57●Compliance in the elderly 58●Effect of drugs on some major organ systems in the elderly 58●Practical aspects of prescribing for the elderly 60●Research 60CHAPTER 11DRUGS IN THE ELDERLYINTRODUCTIONThe proportion of elderly people in the population is increas-ing steadily in economically developed countries. The elderlyare subject to a variety of complaints, many of which arechronic and incapacitating, and so they receive a great deal ofdrug treatment. There is a growing evidence base for the use ofdrugs in elderly patients, with important implications for pre-scribing of many important classes of drugs, including statins,β-adrenoceptor antagonists, thrombolytics, ACE inhibitors,angiotensin receptor blockers, vitamin D and bisphosphonates(see reviews by Mangoni and Jackson, 2006). Adverse drugreactions and drug interactions become more common withincreasing age. In one study, 11.8% of patients aged 41–50years experienced adverse reactions to drugs, but thisincreased to 25% in patients over 80 years of age. There areseveral reasons for this.1. Elderly people take more drugs. In one survey in generalpractice, 87% of patients over 75 years of age were onregular drug therapy, with 34% taking three to fourdifferent drugs daily. The most commonly prescribeddrugs were diuretics (34% of patients), analgesics (27%),tranquillizers and antidepressants (24%), hypnotics (22%)and digoxin (20%). All of these are associated with a highincidence of important adverse effects.2. Drug elimination becomes less efficient with increasingage, leading to drug accumulation during chronic dosing.3. Homeostatic mechanisms become less effective withadvancing age, so individuals are less able to compensatefor adverse effects, such as unsteadiness or posturalhypotension.4. The central nervous system becomes more sensitive to theactions of sedative drugs.5. Increasing age produces changes in the immune responsethat can cause an increased liability to allergic reactions.6. Impaired cognition combined with relatively complexdose regimens may lead to inadvertent overdose.PHARMACOKINETIC CHANGESABSORPTIONAbsorption of carbohydrates and of several nutrients, includ-ing iron, calcium and thiamine, is reduced in elderly people.Lipid-soluble drugs are absorbed by simple diffusion downthe concentration gradient (Chapter 3), and this is notimpaired by age. Intestinal blood flow is reduced by up to 50%in the elderly. However, age per se does not affect drugabsorption to a large extent (Figure 11.1).DrugAbsorption↔Metabolism↓Excretion↓↓Renal blood flow↓GFR↓Concentration offat-soluble drugs↑Concentration ofwater-soluble drugs↓Weight↓Lean body mass↑Fat↑Gastric motility↓Intestinal blood flow↓Hepatic blood flowFigure 11.1: Pharmacokinetic changes with age.
PHARMACODYNAMIC CHANGES 57DISTRIBUTIONAgeing is associated with loss of lean body mass, and with anincreased ratio of fat to muscle and body water. This enlargesthe volume of distribution of fat-soluble drugs, such asdiazepam and lidocaine, whereas the distribution of polardrugs such as digoxin is reduced compared to younger adults.Changes in plasma proteins also occur with ageing, especiallyif associated with chronic disease and malnutrition, with a fallin albumin and a rise in gamma-globulin concentrations.HEPATIC METABOLISMThere is a decrease in the hepatic clearance of some but not alldrugs with advancing age. A prolonged plasma half-life (Figure11.2), can be the result either of reduced clearance or of increasedapparent volume of distribution. Ageing reduces metabolism ofsome drugs (e.g. benzodiazepines) as evidenced by reducedhepatic clearance. The reduced clearance of benzodiazepineshas important clinical consequences, as does the long half-life ofseveral active metabolites (Chapter 18). Slow accumulation maylead to adverse effects whose onset may occur days or weeksafter initiating therapy. Consequently, confusion or memoryimpairment may be falsely attributed to ageing rather than toadverse drug effects.RENAL EXCRETIONThe most important cause of drug accumulation in the elderlyis declining renal function. Many healthy elderly individualshave a glomerular filtration rate (GFR) 50 mL/min. Althoughglomerular filtration rate declines with age, this is not necessar-ily reflected by serum creatinine, which can remain within therange defined as ‘normal’ for a younger adult populationdespite a marked decline in renal function. This is related to thelower endogenous production of creatinine in the elderly sec-ondary to their reduced muscle mass. Under-recognition ofrenal impairment in the elderly is lessened by the routinereporting by many laboratories of an estimated GFR (eGFR)based on age, sex and serum creatinine concentration andreported in units normalized to 1.73 m2body surface area(mL/min/1.73 m2). When estimating doses of nephrotoxicdrugs, it is important to remember that the drug eliminationdepends on the absolute GFR (in mL/min) rather than that nor-malized to an ideal body surface area (in mL/min/1.73 m2),and to estimate this if necessary using a nomogram (seeChapter 7) that incorporates height and weight, as well as age,sex and creatinine.Examples of drugs which may require reduced dosage inthe elderly secondary to reduced renal excretion and/orhepatic clearance are listed in Table 11.1.The principal age-related changes in pharmacokinetics aresummarized in Figure 11.1.Key points1201008060402001020304050Age (years)60 70 80 100Diazepam t1/2 (h)Figure 11.2: Relationship between diazepam half-life and age in33 normal individuals. Non-smokers, °; smokers, •. (Redrawnwith permission from Klotz U et al. Journal of ClinicalInvestigation 1975; 55: 347.)Key pointsPharmacokinetic changes in the elderly include:• Absorption of iron, calcium and thiamine is reduced.• There is an increased volume of distribution of fat-soluble drugs (e.g. diazepam).• There is a decreased volume of distribution of polardrugs (e.g. digoxin).• There is reduced hepatic clearance of long half-lifebenzodiazepines.• Declining renal function is the most important cause ofdrug accumulation.PHARMACODYNAMIC CHANGESEvidence that the elderly are intrinsically more sensitive todrugs than the young is scarce. However, the sensitivity of theelderly to benzodiazepines as measured by psychometric testsis increased, and their effects last longer than in the young. It iscommon clinical experience that benzodiazepines given to theelderly at hypnotic doses used for the young can produce pro-longed daytime confusion even after single doses. The inci-dence of confusion associated with cimetidine is increased inthe elderly. Other drugs may expose physiological defects thatare a normal concomitant of ageing. Postural hypotension canoccur in healthy elderly people, and the incidence of posturalhypotension from drugs such as phenothiazines, β-adrenoceptorTable 11.1 : Examples of drugs requiring dose adjustment in the elderlyAminoglycosides (e.g. gentamicin)AtenololCimetidineDiazepamDigoxinNon-steroidal anti-inflammatory drugsOral hypoglycaemic agentsWarfarin
COMPLIANCE IN THE ELDERLYIncomplete compliance is extremely common in elderly people.This is commonly due to a failure of memory or to not under-standing how the drug should be taken. In addition, manypatients store previously prescribed drugs in the medicinecupboard which they take from time to time. It is thereforeessential that the drug regimen is kept as simple as possibleand explained carefully. There is scope for improved methodsof packaging to reduce over- or under-dosage. Multiple drugregimens are confusing and increase the risk of adverse inter-actions (see Chapter 13).EFFECT OF DRUGS ON SOME MAJORORGAN SYSTEMS IN THE ELDERLYCENTRAL NERVOUS SYSTEMCerebral function in old people is easily disturbed, resulting in disorientation and confusion. Drugs are one of the factorsthat contribute to this state; sedatives and hypnotics can easilyprecipitate a loss of awareness and clouding of consciousness.NIGHT SEDATIONThe elderly do not sleep as well as the young. They sleep for ashorter time, their sleep is more likely to be broken and they aremore easily aroused. This is quite normal, and old people shouldnot have the expectations of the young as far as sleep is con-cerned. Before hypnotics are commenced, other possible factorsshould be considered and treated if possible. These include:1. pain, which may be due to such causes as arthritis;2. constipation – the discomfort of a loaded rectum;3. urinary frequency;4. depression;5. anxiety;58 DRUGS IN THE ELDERLY6. left ventricular failure;7. dementia;8. nocturnal xanthine alkaloids, e.g. caffeine in tea,theophylline.A little more exercise may help, and ‘catnapping’ in the dayreduced to a minimum and regularized (as in Mediterranencultures).The prescription of hypnotics (see Chapter 18) should beminimized and restricted to short-term use.ANTIDEPRESSANTSAlthough depression is common in old age and may indeedneed drug treatment, this is not without risk. Tricyclic anti-depressants (see Chapter 20) can cause constipation, urinaryretention and glaucoma (due to their muscarinic blockingaction which is less marked in the case of lofepramine thanother drugs of this class), and also drowsiness, confusion, pos-tural hypotension and cardiac dysrhythmias. Tricyclic antide-pressants can produce worthwhile remissions of depressionbut should be started at very low dosage.Selective 5-hydroxytryptamine reuptake inhibitors (e.g.fluoxetine) are as effective as the tricyclics and have a distinctside-effect profile (see chapter 20). They are generally well tolerated by the elderly, although hyponatraemia has beenreported more frequently than with other antidepressants.ANTI-PARKINSONIAN DRUGSThe anticholinergic group of anti-parkinsonian drugs (e.g. trihexyphenidyl, orphenadrine) commonly cause side effectsin the elderly. Urinary retention is common in men. Glaucomamay be precipitated or aggravated and confusion may occurwith quite small doses. Levodopa combined with a peripheraldopa decarboxylase inhibitor such as carbidopa can be effec-tive, but it is particularly important to start with a small dose,which can be increased gradually as needed. In patients withdementia, the use of antimuscarinics, levodopa or amantidinemay produce adverse cerebral stimulation and/or hallucin-ations, leading to decompensation of cerebral functioning,with excitement and inability to cope.CARDIOVASCULAR SYSTEMHYPERTENSIONThere is excellent evidence that treating hypertension in theelderly reduces both morbidity and mortality. The agents used(starting with a C or D drug) are described in Chapter 28. It isimportant to start with a low dose and monitor carefully.Some adverse effects (e.g. hyponatraemia from diuretics) aremuch more common in the elderly, who are also much morelikely to suffer severe consequences, such as falls/fracturesfrom common effects like postural hypotension. Alpha-blockers in particular should be used as little as possible.Methyldopa might be expected to be problematic in this agegroup but was in fact surprisingly well tolerated when used asadd-on therapy in a trial by the European Working Party onHypertension in the Elderly (EWPHE).Key pointsPharmacodynamic changes in the elderly include:• increased sensitivity to central nervous system (CNS)effects (e.g. benzodiazepines, cimetidine);• increased incidence of postural hypotension (e.g.phenothiazines, beta-blockers, tricyclic antidepressants,diuretics);• reduced clotting factor synthesis, reduced warfarin foranticoagulation;• increased toxicity from NSAIDs;• increased incidence of allergic reactions to drugs.antagonists, tricyclic antidepressants and diuretics is increasedin elderly patients. The QT interval is longer in the elderly,which may predispose to drug-induced ventricular tachy-dysrhythmias. Clotting factor synthesis by the liver is reducedin the elderly, and old people often require lower warfarindoses for effective anticoagulation than younger adults.
DIGOXINDigoxin toxicity is common in the elderly because of decreasedrenal elimination and reduced apparent volume of distribution.Confusion, nausea and vomiting, altered vision and an acuteabdominal syndrome resembling mesenteric artery obstructionare all more common features of digoxin toxicity in the elderlythan in the young. Hypokalaemia due to decreased potassiumintake (potassium-rich foods are often expensive), faulty homeo-static mechanisms resulting in increased renal loss and the con-comitant use of diuretics is more common in the elderly, and isa contributory factor in some patients. Digoxin is sometimesprescribed when there is no indication for it (e.g. for an irregu-lar pulse which is due to multiple ectopic beats rather thanatrial fibrillation). At other times, the indications for initiation oftreatment are correct but the situation is never reviewed. In oneseries of geriatric patients on digoxin, the drug was withdrawnin 78% of cases without detrimental effects.DIURETICSDiuretics are more likely to cause adverse effects (e.g. posturalhypotension, glucose intolerance and electrolyte disturbances)in elderly patients. Too vigorous a diuresis may result in urin-ary retention in an old man with an enlarged prostate, andnecessitate bladder catheterization with its attendant risks.Brisk diuresis in patients with mental impairment or reducedmobility can result in incontinence. For many patients, a thia-zide diuretic, such as bendroflumethiazide, is adequate. Loopdiuretics, such as furosemide, should be used in acute heartfailure or in the lowest effective dose for maintenance treatmentof chronic heart failure. Clinically important hypokalaemia isuncommon with low doses of diuretics, but plasma potassiumshould be checked after starting treatment. If clinically importanthypokalaemia develops, a thiazide plus potassium-retainingdiuretic (amiloride or triamterene) can be considered, but thereis a risk of hyperkalaemia due to renal impairment, especially if an ACE inhibitor and/or angiotensin receptor antagonist and aldosterone antagonist are given together with the diureticfor hypertension or heart failure. Thiazide-induced gout andglucose intolerance are important side effects.ISCHAEMIC HEART DISEASEThis is covered in Chapter 29.ANGIOTENSIN CONVERTING ENZYME INHIBITORS(ACEI) AND ANGIOTENSIN RECEPTOR BLOCKERS (ARB)These drugs plays an important part in the treatment of chronicheart failure, as well as hypertension (see Chapters 28 and 31),and are effective and usually well tolerated in the elderly. However, hypotension, hyperkalaemia and renal failure aremore common in this age group. The possibility of atheroma-tous renal artery stenosis should be borne in mind and serumcreatinine levels checked before and after starting treatment.Potassium-retaining diuretics should be co-administered onlywith extreme caution, because of the reduced GFR and plasmapotassium levels monitored. Despite differences in their phar-macology, ACEI and ARB appear similar in efficacy, but ARB do not cause the dry cough that is common with ACEI. TheEFFECT OF DRUGS ON SOME MAJOR ORGAN SYSTEMS IN THE ELDERLY 59question of whether co-administration of ACEI with ARB hasmuch to add remains controversial; in elderly patients withreduced GFR, the safety of such combined therapy is an impor-tant consideration.ORAL HYPOGLYCAEMIC AGENTSDiabetes is common in the elderly and many patients aretreated with oral hypoglycaemic drugs (see Chapter 37). It isbest for elderly patients to be managed with diet if at all possi-ble. In obese elderly diabetics who remain symptomatic ondiet, metformin should be considered, but coexisting renal,heart or lung disease may preclude its use. Short-actingsulphonylureas (e.g. gliclazide) are preferred to longer-actingdrugs because of the risk of hypoglycaemia: chlorpropamide(half-life 36 hours) can cause prolonged hypoglycaemia and isspecifically contraindicated in this age group, glibenclamideshould also be avoided. Insulin may be needed, but impairedvisual and cognitive skills must be considered on an individualbasis, and the potential need for dose reduction with advanc-ing age and progressive renal impairment taken into account.ANTIBIOTICSThe decline in renal function must be borne in mind when anantibiotic that is renally excreted is prescribed, especially if it isnephrotoxic (e.g. an aminoglycoside or tetracycline). Appendix 3of the British National Formulary is an invaluable practical guide.Over-prescription of antibiotics is a threat to all age groups, but especially in the elderly. Broad-spectrum drugs includingcephalosporins and other beta-lactams, and fluoroquinones arecommon precursors of Clostridium difficile infection which has ahigh mortality rate in the elderly. Amoxicillin is the most com-mon cause of drug rash in the elderly. Flucloxacillin inducedcholestatic jaundice and hepatitis is more common in the elderly.Case historyAn 80-year-old retired publican was referred with ‘congest-ive cardiac failure and acute retention of urine’. His wifesaid his symptoms of ankle swelling and breathlessness hadgradually increased over a period of six months despite theGP doubling the water tablet (co-amilozide) which he wastaking for high blood pressure. Over the previous week hehad become mildly confused and restless at night, forwhich the GP had prescribed chlorpromazine. His othermedication included ketoprofen for osteoarthritis and fre-quent magnesium trisilicate mixture for indigestion. Hehad been getting up nearly ten times most nights for a yearto pass urine. During the day, he frequently passed smallamounts of urine. Over the previous 24 hours, he had beenunable to pass urine. His wife thought most of his problemswere due to the fact that he drank two pints of beer eachday since his retirement seven years previously.On physical examination he was clinically anaemic, butnot cyanosed. Findings were consistent with congestivecardiac failure. His bladder was palpable up to his umbili-cus. Rectal examination revealed an enlarged, symmetricalprostate and black tarry faeces. Fundoscopy revealed agrade II hypertensive retinopathy.
60 DRUGS IN THE ELDERLY4.Use the fewest possible number of drugs the patient needs.5. Consider the potential for drug interactions and co-morbidity on drug response.6. Drugs should seldom be used to treat symptoms withoutfirst discovering the cause of the symptoms (i.e. firstdiagnosis, then treatment).7. Drugs should not be withheld because of old age, but itshould be remembered that there is no cure for old ageeither.8. A drug should not be continued if it is no longer necessary.9. Do not use a drug if the symptoms it causes are worsethan those it is intended to relieve.10 . It is seldom sensible to treat the side effects of one drugby prescribing another.In the elderly, it is often important to pay attention to mat-ters such as the formulation of the drug to be used – many oldpeople tolerate elixirs and liquid medicines better than tabletsor capsules. Supervision of drug taking may be necessary, asan elderly person with a serious physical or mental disabilitycannot be expected to comply with any but the simplest drugregimen. Containers require especially clear labelling, andshould be easy to open – child-proof containers are often alsograndparent-proof!RESEARCHDespite their disproportionate consumption of medicines, theelderly are often under-represented in clinical trials. This mayresult in the data being extrapolated to an elderly populationinappropriately, or the exclusion of elderly patients from newtreatments from which they might benefit. It is essential that,both during a drug’s development and after it has been licensed,subgroup analysis of elderly populations is carefully examinedboth for efficacy and for predisposition to adverse effects.Initial laboratory results revealed that the patient hadacute on chronic renal failure, dangerously high potassiumlevels (7.6 mmol/L) and anaemia (Hb 7.4 g/dL). Emergencytreatment included calcium chloride, dextrose and insulin,urinary catheterization, furosemide and haemodialysis.Gastroscopy revealed a bleeding gastric ulcer. The patientwas discharged two weeks later, when he was symptomat-ically well. His discharge medication consisted of regulardoxazosin and ranitidine, and paracetamol as required.QuestionDescribe how each of this patient’s drugs prescribed beforeadmission may have contributed to his clinical condition.AnswerCo-amilozide – hyperkalaemia: amiloride, exacerbation ofprostatic symptoms: thiazideChlorpromazine – urinary retentionKetoprofen – gastric ulcer, antagonism of thiazide diuretic,salt retention, possibly interstitial nephritisMagnesium trisilicate mixture – additional sodium load(6 mmol Na/10 mL).CommentIatrogenic disease due to multiple drug therapy is commonin the elderly. The use of amiloride in renal impairmentleads to hyperkalaemia. This patient’s confusion and rest-lessness were most probably related to his renal failure.Chlorpromazine may mask some of the symptoms/signs anddelay treatment of the reversible organic disease. The anal-gesic of choice in osteoarthritis is paracetamol, due to itsmuch better tolerance than NSAID. The sodium content ofsome antacids can adversely affect cardiac and renal failure.NON-STEROIDAL ANTI-INFLAMMATORY DRUGSThe elderly are particularly susceptible to non-steroidal anti-inflammatory drug (NSAID)-induced peptic ulceration, gastro-intestinal irritation and fluid retention. An NSAID is frequentlyprescribed inappropriately for osteoarthritis before physicaland functional interventions and oral paracetamol have beenadequately utilized. If an NSAID is required as adjunctivetherapy, the lowest effective dose should be used. Ibuprofenis probably the NSAID of choice in terms of minimizing gas-tro-intestinal side effects. A proton pump inhibitor should beconsidered as prophylaxis against upper gastro-intestinalcomplications in those most at risk.PRACTICAL ASPECTS OF PRESCRIBINGFOR THE ELDERLYImproper prescription of drugs is a common cause of morbid-ity in elderly people. Common-sense rules for prescribing donot apply only to the elderly, but are especially important inthis vulnerable group.1. Take a full drug history (see Chapter 1), which shouldinclude any adverse reactions and use of over-the-counterdrugs.2. Know the pharmacological action of the drug employed.3. Use the lowest effective dose.Case historyA previously mentally alert and well-orientated 90-year-oldwoman became acutely confused two nights after hospitaladmission for bronchial asthma which, on the basis of peakflow and blood gases, had responded well to inhaled salbu-tamol and oral prednisolone. Her other medication wascimetidine (for dyspepsia), digoxin (for an isolated episodeof atrial fibrillation two years earlier) and nitrazepam (fornight sedation).QuestionWhich drugs may be related to the acute confusion?AnswerPrednisolone, cimetidine, digoxin and nitrazepam.CommentIf an H2-antagonist is necessary, ranitidine is preferred in theelderly. It is likely that the patient no longer requires digoxin(which accumulates in the elderly). Benzodiazepines shouldnot be used for sedation in elderly (or young) asthmatics.They may also accumulate in the elderly. The elderly tend tobe more sensitive to adverse drug effects on the central ner-vous system (CNS).
RESEARCH 61FURTHER READINGDhesi JK, Allain TJ, Mangoni AA, Jackson SHD. The implications of agrowing evidence base for drug use in elderly patients. Part 4.Vitamin D and bisphosphonates for fractures and osteoporosis.British Journal of Clinical Pharmacology 2006; 61: 520–8.Hanratty CG, McGlinchey P, Johnston GD, Passmore AP. Differentialpharmacokinetics of digoxin in elderly patients. Drugs and Aging2000; 17: 353–62.Mangoni AA, Jackson SHD. The implications of a growing evidencebase for drug use in elderly patients. Part 1. Statins for primaryand secondary cardiovascular prevention. British Journal of ClinicalPharmacology 2006; 61: 494–501.Mangoni AA, Jackson SHD. The implications of a growing evidencebase for drug use in elderly patients. Part 2. ACE inhibitors andangiotensin receptor blockers in heart failure and high cardiovas-cular risk patients. British Journal of Clinical Pharmacology 2006; 61:502–12.Mangoni AA, Jackson SHD. The implications of a growing evidencebase for drug use in elderly patients. Part 3. β-adrenoceptor block-ers in heart failure and thrombolytics in acute myocardial infarc-tion. British Journal of Clinical Pharmacology 2006; 61: 513–20.Sproule BA, Hardy BG, Shulman KI. Differential pharmacokinetics inelderly patients. Drugs and Aging 2000; 16: 165–77.
62●Identification of the drug at fault 63●Adverse drug reaction monitoring/surveillance(pharmacovigilance) 63●Allergic adverse drug reactions 66●Prevention of allergic drug reactions 67●Examples of allergic and other adversedrug reactions 68CHAPTER 12ADVERSE DRUG REACTIONSINTRODUCTIONAdverse drug reactions are unwanted effects caused by nor-mal therapeutic doses. Drugs are great mimics of disease, and adverse drug reactions present with diverse clinical signs and symptoms. The classification proposed by Rawlinsand Thompson (1977) divides reactions into type A and type B(Table 12.1).Type A reactions, which constitute approximately 80% ofadverse drug reactions, are usually a consequence of the drug’sprimary pharmacological effect (e.g. bleeding from warfarin)or a low therapeutic index (e.g. nausea from digoxin), and theyare therefore predictable. They are dose-related and usuallymild, although they may be serious or even fatal (e.g. intracra-nial bleeding from warfarin). Such reactions are usually due toinappropriate dosage, especially when drug elimination isimpaired. The term ‘side effects’ is often applied to minor typeA reactions.Type B (‘idiosyncratic’) reactions are not predictable fromthe drug’s main pharmacological action, are not dose-relatedand are severe, with a considerable mortality. The underlyingpathophysiology of type B reactions is poorly if at all under-stood, and often has a genetic or immunological basis. Type Breactions occur infrequently (1:1000–1:10 000 treated subjectsbeing typical).Adverse drug reactions due to specific drug–drug inter-actions are considered in Chapter 13. Three further minor cat-egories of adverse drug reaction have been proposed:1. type C – continuous reactions due to long-term drug use(e.g. neuroleptic-related tardive dyskinesia or analgesicnephropathy);2. type D – delayed reactions (e.g. alkylating agents leading to carcinogenesis, or retinoid-associatedteratogenesis);3. type E end-of-use reactions, such as adrenocorticalinsufficiency following withdrawal of glucocorticosteroids,or withdrawal syndromes following discontinuation oftreatment with benzodiazepines or β-adrenoceptorantagonists.In the UK there are between 30 000 and 40 000 medicinalproducts available directly or on prescription. Surveys sug-gest that approximately 80% of adults take some kind of medi-cation during any two-week period. Exposure to drugs in thepopulation is thus substantial, and the incidence of adversereactions must be viewed in this context. Type A reactions arereported to be responsible for 2–3% of consultations in generalpractice. In a recent prospective analysis of 18 820 hospitaladmissions by Pirmohamed et al. (2004), 1225 were related toan adverse drug reaction (prevalence 6.8%), with the adversedrug reaction leading directly to admission in 80% of cases.Median bed stay was eight days, accounting for 4% of hospi-tal bed capacity. The projected annual cost to the NHS is £466 million. Overall fatality was 0.15%. Most reactions were eitherdefinitely or probably avoidable. Adverse drug reactions aremost frequent and severe in the elderly, in neonates, women,patients with hepatic or renal impairment, and individualswith a history of previous adverse drug reactions. Such reac-tions often occur early in therapy (during the first one to tendays). Drugs most commonly implicated include low-doseaspirin (antiplatelet agents), diuretics, warfarin and NSAIDs.A systematic review by Howard et al. (2006) of preventableadverse drug reactions which caused hospitalization, impli-cated the same major drug classes.Table 12.1: Some examples of type A and type B reactions.Drug Type A Type BChlorpromazine Sedation Cholestatic jaundiceNaproxen Gastro-intestinal AgranulocytosishaemorrhagePhenytoin Ataxia Hepatitis,lymphadenopathyThiazides Hypokalaemia ThrombocytopeniaQuinine Tinnitus ThrombocytopeniaWarfarin Bleeding Breast necrosis
ADVERSE DRUG REACTION MONITORING/SURVEILLANCE (PHARMACOVIGILANCE) 63Factors involved in the aetiology of adverse drug reactionscan be classified as shown in Table 12.2.IDENTIFICATION OF THE DRUG AT FAULTIt is often difficult to decide whether a clinical event is drugrelated, and even when this is probable, it may be difficult todetermine which drug is responsible, as patients are often tak-ing multiple drugs. One or more of several possible approachesmay be appropriate.1. A careful drug history is essential. The followingconsiderations should be made to assess causality of theeffect to the drug: did the clinical event and the time-course of its development fit with the duration of suspecteddrug treatment and known adverse drug effects? Did theadverse effect reverse upon drug withdrawal and, uponrechallenge with the drug, reappear? Were other possiblecauses reasonably excluded? A patient’s drug history maynot always be conclusive because, although allergy to adrug implies previous exposure, the antigen may haveoccurred in foods (e.g. antibiotics are often fed to livestockand drug residues remain in the flesh), in drug mixturesor in some casual manner.2. Provocation testing. This involves giving a very smallamount of the suspected drug and seeing whether areaction ensues, e.g. skin testing, where a drug is appliedas a patch, or is pricked or scratched into the skin orinjected intradermally. Unfortunately, prick and scratchtesting is less useful for assessing the systemic reaction todrugs than it is for the more usual atopic antigens (e.g.pollens), and both false-positive and false-negative resultscan occur. Patch testing is safe, and is useful for thediagnosis of contact sensitivity, but does not reflectsystemic reactions and may itself cause allergy. Provocationtests should only be undertaken under expert guidance,after obtaining informed consent, and with resuscitationfacilities available.3. Serological testing and lymphocytes testing. Serologicaltesting is rarely helpful, circulating antibodies to the drugdo not mean that they are necessarily the cause of thesymptoms. The demonstration of transformation occurringwhen the patient’s lymphocytes are exposed to a drug ex vivo suggests that the patient’s T-lymphocytes aresensitized to the drug. In this type of reaction, the haptenitself will often provoke lymphocyte transformation, aswell as the conjugate.4. The best approach in patients on multiple drug therapy isto stop all potentially causal drugs and reintroduce themone by one until the drug at fault is discovered. This shouldonly be done if the reaction is not serious, or if the drug is essential and no chemically unrelated alternative isavailable. All drug allergies should be recorded in the casenotes and the patient informed of the risks involved intaking the drug again.Key points• Type A reaction – an extension of the pharmacology ofthe drug, dose related, and accounts for most adversereactions (e.g. β-adrenoreceptor antagonist-inducedbradycardia or AV block).• Type B reaction – idiosyncratic reaction to the drug, notdose related, rare but severe (e.g. chloramphenicol-induced aplastic anaemia).• Other types of drug reaction (much rarer):– type C reaction – continuous reactions due to long-term use: analgesic nephropathy;– type D reaction – delayed reactions ofcarcinogenesis or teratogenesis;– type E reaction – drug withdrawal reactions (e.g.benzodiazepines).ADVERSE DRUG REACTION MONITORING/SURVEILLANCE (PHARMACOVIGILANCE)The evaluation of drug safety is complex, and there are manymethods for monitoring adverse drug reactions. Each of thesehas its own advantages and shortcomings, and no single Table 12.2: Factors involved in adverse drug reactions.Intrinsic ExtrinsicPatient factorsAge – neonatal, infant and elderly Environment – sunSex – hormonal environment Xenobiotics (e.g. drugs,Genetic abnormalities (e.g. herbicides)enzyme or receptor Malnutritionpolymorphisms)Previous adverse drug reactions,allergy, atopyPresence of organ dysfunction –diseasePersonality and habits –adherence (compliance),alcoholic, drug addict,nicotinePrescriber factorsIncorrect drug or drug combinationIncorrect route of administrationIncorrect doseIncorrect duration of therapyDrug factorsDrug–drug interactions (seeChapter 13)Pharmaceutical – batch problems,shelf-life, incorrect dispensing
64 ADVERSE DRUG REACTIONSsystem can offer the 100% accuracy that current public opinionexpects. The ideal method would identify adverse drug reactionswith a high degree of sensitivity and specificity and respondrapidly. It would detect rare but severe adverse drug reactions,but would not be overwhelmed by common ones, the incidenceof which it would quantify together with predisposing factors.Continued surveillance is mandatory after a new drug hasbeen marketed, as it is inevitable that the preliminary testingof medicines in humans during drug development, althoughexcluding many ill effects, cannot identify uncommon adverseeffects. A variety of early detection systems have been intro-duced to identify adverse drug reactions as swiftly as possible.PHASE I/II/III TRIALSEarly (phase I/II) trials (Chapter 15) are important for assess-ing the tolerability and dose–response relationship of newtherapeutic agents. However, these studies are, by design,very insensitive at detecting adverse reactions because theyare performed on relatively few subjects (perhaps 200–300).This is illustrated by the failure to detect the serious toxicity of several drugs (e.g. benoxaprofen, cerivastatin, felbamate,dexfenfluramine and fenfluramine, rofecoxib, temofloxacin,troglitazone) before marketing. However, phase III clinical trials can establish the incidence of common adverse reactionsand relate this to therapeutic benefit. Analysis of the reasonsgiven for dropping out of phase III trials is particularly valu-able in establishing whether common events, such asheadache, constipation, lethargy or male sexual dysfunctionare truly drug related. The Medical Research Council MildHypertension Study unexpectedly identified impotence asmore commonly associated with thiazide diuretics than withplacebo or β-adrenoceptor antagonist therapy. Table 12.3 illustrates how difficult it is to detect adverse drug reactionswith 95% confidence, even when there is no background inci-dence and the diagnostic accuracy is 100%. This ‘easiest-case’scenario approximates to the actual situation with thalido-mide teratogenicity: spontaneous phocomelia is almostunknown, and the condition is almost unmistakable. It issobering to consider that an estimated 10 000 malformedbabies were born world-wide before thalidomide was with-drawn. Regulatory authorities may act after three or moredocumented events.The problem of adverse drug reaction recognition is muchgreater if the reaction resembles spontaneous disease in thepopulation, such that physicians are unlikely to attribute thereaction to drug exposure: the numbers of patients that mustthen be exposed to enable such reactions to be detected aregreater than those quoted in Table 12.3, probably by severalorders of magnitude.YELLOW CARD SCHEME AND POST-MARKETING (PHASE IV) SURVEILLANCEUntoward effects that have not been detected in clinical trialsbecome apparent when the drug is used on a wider scale. Casereports, which may stimulate further reports, remain the mostsensitive means of detecting rare but serious and unusualadverse effects. In the UK, a Register of Adverse Reactionswas started in 1964. Currently, the Medicines and Healthcareproducts Regulatory Agency (MHRA) operates a system ofspontaneous reporting on prepaid yellow postcards. Doctors,dentists, pharmacists, nurse practitioners and (most recently)patients are encouraged to report adverse events whether actu-ally or potentially causally drug-related. Analogous schemes areemployed in other countries. The yellow card scheme consistsof three stages:1. data collection;2. analysis;3. feedback.Such surveillance methods are useful, but under-reporting is amajor limitation. Probably fewer than 10% of appropriateadverse reactions are reported. This may be due partly to con-fusion about what events to report, partly to difficulty in rec-ognizing the possible relationship of a drug to an adverseevent – especially when the patient has been taking severaldrugs, and partly to ignorance or laziness on the part of poten-tial reporters. A further problem is that, as explained above, ifa drug increases the incidence of a common disorder (e.g.ischaemic heart disease), the change in incidence must be verylarge to be detectable. This is compounded when there is adelay between starting the drug and occurrence of the event(e.g. cardiovascular thrombotic events including myocardialinfarction following initiation of rofecoxib therapy). Doctorsare inefficient at detecting such adverse reactions to drugs,and those reactions that are reported are in general the obvi-ous or previously described and well-known ones. Initiativesare in progress to attempt to improve this situation by involve-ment of trained clinical pharmacologists and pharmacists inand outside hospitals.The Committee on Safety of Medicines (CSM), now part ofMHRA, introduced a system of high vigilance for newly mar-keted drugs. For its first two years on the general market, anynewly marketed drug has a black triangle on its data sheet andagainst its entry in the British National Formulary. This con-veys to prescribers that any unexpected event should bereported by the yellow card system. The pharmaceutical com-pany is also responsible for obtaining accurate reports on allpatients treated up to an agreed number. This scheme wassuccessful in the case of benoxaprofen, an anti-inflammatoryTable 12.3: Numbers of subjects that would need to be exposed in order todetect adverse drug reactionsExpected frequency Approximate number of patientsof the adverse effect required to be exposedFor one event For three events1 in 100 300 6501 in 1000 3000 65001 in 10 000 30 000 65 000
ADVERSE DRUG REACTION MONITORING/SURVEILLANCE (PHARMACOVIGILANCE) 65analgesic. Following its release, there were spontaneous reportsto the CSM of photosensitivity and onycholysis. Further reportsappeared in the elderly, in whom its half-life is prolonged, ofcholestatic jaundice and hepatorenal failure, which was fatalin eight cases. Benoxaprofen was subsequently taken off themarket when 3500 adverse drug reaction reports were receivedwith 61 fatalities. The yellow card/black triangle scheme wasalso instrumental in the early identification of urticaria andcough as adverse effects of angiotensin-converting enzymeinhibitors. Although potentially the population under studyby this system consists of all the patients using a drug, in factunder-reporting yields a population that is not uniformlysampled. Such data can be unrepresentative and difficult towork with statistically, contributing to the paucity of accurateincidence data for adverse drug reactions.Systems such as the yellow card scheme (e.g. FDAMedWatchin the USA) are relatively inexpensive and easy to manage,and facilitate ongoing monitoring of all drugs, all consumersand all types of adverse reaction. Reports from the drug regu-latory bodies of 22 countries are collated by the World HealthOrganization (WHO) Unit of Drug Evaluation and Monitoringin Geneva. Rapid access to reports from other countries shouldbe of great value in detecting rare adverse reactions, althoughthe same reservations apply to this register as apply tonational systems. In addition, this database could reveal geo-graphical differences in the pattern of untoward drug effects.CASE–CONTROL STUDIESA very large number of patients have to be monitored to detecta rare type B adverse effect. An alternative approach is to iden-tify patients with a disorder which it is postulated could becaused by an adverse reaction to a drug, and to compare the fre-quency of exposure to possible aetiological agents with a con-trol group. A prior suspicion (hypothesis) must exist to promptthe setting up of such a study – examples are the possible con-nection between irradiation or environmental pollution andcertain malignancies, especially where they are observed inclusters. Artefacts can occur as a result of unrecognized biasfrom faulty selection of patients and controls, and the approachremains controversial among epidemiologists, public healthphysicians and statisticians. Despite this, there is really no prac-ticable alternative for investigating a biologically plausiblehypothesis relating to a disease which is so uncommon that itis unlikely to be represented even in large trial or cohort popu-lations. This methodology has had notable successes: the associ-ation of stilboestrol with vaginal adenocarcinoma, gatifloxacinwith hypo- and hyperglycaemia, and salmeterol or fenoteroluse with increased fatality in asthmatics.INTENSIVE MONITORINGSeveral hospital-based intensive monitoring programmes arecurrently in progress. The Aberdeen–Dundee system abstractsdata from some 70 000 hospital admissions each year, storingthese on a computer file before analysis. The BostonCollaborative Drug Surveillance Program (BCDSP), involvingselected hospitals in several countries, is even more compre-hensive. In the BCDSP, all patients admitted to specially desig-nated general wards are included in the analysis. Speciallytrained personnel obtain the following information from hos-pital patients and records:1. background information (i.e. age, weight, height, etc.);2. medical history;3. drug exposure;4. side effects;5. outcome of treatment and changes in laboratory testsduring hospital admission.A unique feature of comprehensive drug-monitoring sys-tems lies in their potential to follow up and investigate adversereactions suggested by less sophisticated detection systems, orby isolated case reports in medical journals. Furthermore, thefrequency of side effects can be determined more cheaply thanby a specially mounted trial to investigate a simple adverseeffect. Thus, for example, the risk of developing a rash withampicillin was found to be around 7% both by clinical trialand by the BCDSP, which can quantify such associationsalmost automatically from data on its files. New adverse reac-tions or drug interactions are sought by multiple correlationanalysis. Thus, when an unexpected relationship arises, suchas the 20% incidence of gastro-intestinal bleeding in severelyill patients treated with ethacrynic acid compared to 4.3%among similar patients treated with other diuretics, this can-not be attributed to bias arising from awareness of the hypoth-esis during data collection, since the data were collectedbefore the hypothesis was proposed. Conversely, there is apossibility of chance associations arising from multiple com-parisons (‘type I’ statistical error), and such associations mustbe reviewed critically before accepting a causal relationship. Itis possible to identify predisposing risk factors. In the associ-ation between ethacrynic acid and gastro-intestinal bleeding,these were female sex, a high blood urea concentration, previ-ous heparin administration and intravenous administration ofthe drug. An important aspect of this type of approach is thatlack of clinically important associations can also be investi-gated. Thus, no significant association between aspirin andrenal disease was found, whereas long-term aspirin consump-tion is associated with a decreased incidence of myocardialinfarction, an association which has been shown to be of thera-peutic importance in randomized clinical trials (Chapter 29).There are plans to extend intensive drug monitoring to coverother areas of medical practice.However, in terms of new but uncommon adverse reac-tions, the numbers of patients undergoing intensive monitor-ing while taking a particular drug will inevitably be too smallfor the effect to be detectable. Such monitoring can thereforeonly provide information about relatively common, early reac-tions to drugs used under hospital conditions. Patients are notin hospital long enough for detection of delayed effects, whichare among the reactions least likely to be recognized as sucheven by an astute clinician.
66 ADVERSE DRUG REACTIONSMONITORING FROM NATIONAL STATISTICSA great deal of information is available from death certificates,hospital discharge diagnoses and similar records. From thesedata, it may be possible to detect a change in disease trendsand relate this to drug therapy. Perhaps the best-known exampleof this is the increased death rate in young asthmatics noted inthe mid-1960s, which was associated with overuse of bron-chodilator inhalers containing non-specific β-adrenoceptoragonists (e.g. adrenaline and/or isoprenaline). Although rel-atively inexpensive, the shortcomings of this method are obvi-ous, particularly in diseases with an appreciable mortality,since large numbers of patients must suffer before the change isdetectable. Data interpretation is particularly difficult whenhospital discharges are used as a source of information, sincedischarge diagnosis is often provisional or incomplete, andmay be revised during follow up.However, they can combine with high molecular weight enti-ties, usually proteins, to form an antigenic hapten conjugate.The factors that determine the development of allergy to adrug are not fully understood. Some drugs (e.g. penicillin)are more likely to cause allergic reactions than others, andtype I (immediate anaphylactic) reactions are more commonin patients with a history of atopy. A correlation between aller-gic reactions involving immunoglobulin E (IgE) and humanleukocyte antigen (HLA) serotypes has been reported, so geneticfactors may also be important. There is some evidence thatdrug allergies are more common in older people, in womenand in those with a previous history of drug reaction. However,this may merely represent increased frequencies of drug expo-sure in these patient groups.TYPES OF ALLERGYDrugs cause a variety of allergic responses (Figure 12.1) andsometimes a single drug can be responsible for more than onetype of allergic response.TYPE I REACTIONSType I reactions are due to the production of reaginic (IgE)antibodies to an antigen (e.g. penicillins and cephalosporins).The antigen binds to surface bound IgE on mast cells causingdegranulation and release of histamine, eicosanoids andcytokines. It commonly occurs in response to a foreign serumor penicillin, but may also occur with streptomycin and somelocal anaesthetics. With penicillin, it is believed that the peni-cilloyl moiety of the penicillin molecule is responsible for theproduction of antibodies. Treatment of anaphylactic shock isdetailed in Chapter 50.TYPE II REACTIONSThese are due to antibodies of class IgG and IgM which, oncontact with antibodies on the surface of cells, bind comple-ment, causing cell lysis (e.g. penicillin, cephalosporins,methyldopa or quinine) causing, for example, Coombs’ posi-tive haemolytic anaemia.TYPE III IMMUNE COMPLEX ARTHUS REACTIONSCirculating immune complexes can produce several clinicalallergic states, including serum sickness and immune complexglomerulonephritis, and a syndrome resembling systemic lupuserythematosus. The onset of serum sickness is delayed for sev-eral days until features develop such as fever, urticaria,arthropathy, lymphadenopathy, proteinuria and eosinophilia.Recovery takes a few days. Examples of causative agentsinclude serum, penicillin, sulfamethoxazole/trimethoprim,streptomycin and propylthiouracil. Amiodarone lung andhydralazine-induced systemic lupus syndrome are also pos-sibly mediated by immune complex-related mechanisms,although these reactions are less well understood.TYPE IV DELAYED HYPERSENSITIVITY REACTIONSType IV reactions are delayed hypersensitivity reactions, theclassical example of which is contact dermatitis (e.g. to topicalKey points• Rare (and often severe) adverse drug events may not bedetected in early drug development but only defined inthe first few years post marketing (phase IV of drugdevelopment).• Be aware of and participate in the MHRA yellow cardsystem for reporting suspected adverse drug reactions.• Use of any recently marketed drug, which is identifiedwith a black triangle on its data sheet or in the BritishNational Formulary, indicates the need to be particularlysuspicious about adverse drug reactions and to reportany suspected adverse drug reaction via the yellow cardsystem.• Constant vigilance by physicians for drug-induceddisease, particularly for new drugs, but also for moreestablished agents, is needed.FEEDBACKThere is no point in collecting vast amounts of data on adverse reactions unless they are analysed and conclusionsreported back to prescribing doctors. In addition to articles inthe medical journals and media, the Current Problems inPharmacovigilance series deals with important and recentlyidentified adverse drug reactions. If an acute and seriousproblem is recognized, doctors will usually receive notifica-tion from the MHRA/Commission on Human Medicines, and often from the pharmaceutical company marketing theproduct.ALLERGIC ADVERSE DRUG REACTIONSImmune mechanisms are involved in a number of adverseeffects caused by drugs (see below and Chapter 50). Thedevelopment of allergy implies previous exposure to the drugor to some closely related substance. Most drugs are of lowmolecular weight (300–500 Da) and thus are not antigenic.
PREVENTION OF ALLERGIC DRUG REACTIONS 67antibiotics, such as penicillin or neomycin). The mechanismhere is that the drug applied to the skin forms an antigenicconjugate with dermal proteins, stimulating formation of sen-sitized T-lymphocytes in the regional lymph nodes, with aresultant rash if the drug is applied again. Drug photosensitiv-ity is due to a photochemical combination between the drug(e.g. amiodarone, chlorpromazine, ciprofloxacin, tetracyc-lines) and dermal protein. Delayed sensitivity can also resultfrom the systemic administration of drugs.vitamin supplements and alternative remedies) is essential.A history of atopy, although not excluding the use ofdrugs, should make one wary.2. Drugs given orally are less likely to cause severe allergicreactions than those given by injection.3. Desensitization (hyposensitization) should only be usedwhen continued use of the drug is essential. It involvesgiving a very small dose of the drug and increasing thedose at regular intervals, sometimes under cover of aglucocorticosteroid and β2-adrenoceptor agonist. Anantihistamine may be added if a drug reaction occurs, andequipment for resuscitation and therapy of anaphylacticshock must be close at hand. It is often successful,although the mechanism by which it is achieved is notfully understood.4. Prophylactic skin testing is not usually practicable, and anegative test does not exclude the possibility of an allergicreaction.Key pointsHow to attempt to define the drug causing the adversedrug reaction:• Attempt to define the likely causality of the effect tothe drug, thinking through the following: Did thereaction and its time-course fit with the duration ofsuspected drug treatment and known adverse drugeffects? Did the adverse effect disappear on drugwithdrawal and, if rechallenged with the drug,reappear? Were other possible causes excluded?• Provocation testing with skin testing – intradermal testsare neither very sensitive nor specific.• Test the patient’s serum for anti-drug antibodies, or testthe reaction of the patient’s lymphocytes in vitro to thedrug and/or drug metabolite if appropriate.• Consider stopping all drugs and reintroducing essentialones sequentially.• Carefully document and highlight the adverse drugreaction and the most likely culprit in the case notes.Key pointsClassification of immune-mediated adverse drug reactions:• Type I – urticaria or anaphylaxis due to the productionof IgE against drug bound to mast cells, leading tomassive release of mast cell mediators locally orsystemically (e.g. ampicillin skin allergy or anaphylaxis).• Type II – IgG and IgM antibodies to drug which, oncontact with antibodies on the cell surface, cause celllysis by complement fixation (e.g. penicillin, haemolyticanaemia; quinidine, thrombocytopenia).• Type III – circulating immune complexes produced bydrug and antibody to drug deposit in organs, causingdrug fever, urticaria, rash, lymphadenopathy,glomerulonephritis, often with eosinophilia (e.g. co-trimoxazole, β-lactams).• Type IV – delayed-type hypersensitivity due to drugforming an antigenic conjugate with dermal proteinsand sensitized T cells reacting to drug, causing a rash(e.g. topical antibiotics).Central immuneapparatusType IV responseType I, II and III responsesHumoral antibodiesTLymphocytesSensitizedlymphocytesCellmembraneDrug or itsmetabolitesDrug or itsmetabolitesDrug(large molecule)Protein AntigenBLymphocytesMacrophagesFigure 12.1: The immune response to drugs.PREVENTION OF ALLERGIC DRUGREACTIONSAlthough it is probably not possible to avoid all allergic drugreactions, the following measures can decrease their incidence:1. Taking a detailed drug history (prescription and over-the-counter drugs, drugs of abuse, nutritional and
68 ADVERSE DRUG REACTIONSEXAMPLES OF ALLERGIC AND OTHERADVERSE DRUG REACTIONSAdverse drug reactions can be manifested in any one or mul-tiple organ systems, and in extraordinarily diverse forms.Specific instances are dealt with throughout this book. Someexamples to illustrate the diversity of adverse drug reactionsare given here.RASHESThese are one of the most common manifestations of drugreactions. A number of immune and non-immune mech-anisms may be involved which produce many different typesof rash ranging from a mild maculopapular rash to a severeerythema multiforme major (Stevens Johnson syndrome;Figures 12.2 and 12.3). Commonly implicated drugs/drugclasses include beta-lactams, sulphonamides and other anti-microbial agents; anti-seizure medications (e.g. phenytoin,carbamazepine); NSAIDs. Some drugs may give rise to directtissue toxicity (e.g. DMPS, used as chelating therapy in patientswith heavy metal poisoning; Figure 12.4, see Chapter 54).LYMPHADENOPATHYLymph-node enlargement can result from taking drugs (e.g.phenytoin). The mechanism is unknown, but allergic factorsmay be involved. The reaction may be confused with a lymph-oma, and the drug history is important in patients with lym-phadenopathy of unknown cause.BLOOD DYSCRASIASThrombocytopenia, anaemia (aplastic, iron deficiency, macro-cytic, haemolytic) and agranulocytosis can all be caused bydrugs.Thrombocytopenia can occur with many drugs, and inmany but not all instances the mechanism is direct suppres-sion of the megakaryocytes rather than immune processes.Drugs that cause thrombocytopenia include:• heparin;• gold salts;• cytotoxic agents (e.g. azathioprine/6-mercaptopurine);• quinidine;• sulphonamides;• thiazides.Haemolytic anaemia can be caused by a number of drugs, and sometimes immune mechanisms are responsible.Glucose-6-phosphate dehydrogenase deficiency (Chapter 14)Figure 12.2: Mouth ulcer as part of Stevens Johnson syndrome asa reaction to phenytoin therapy (see Chapter 22).Figure 12.3: Stevens Johnson syndrome followingcommencement of penicillin therapy (see Chapter 43).Figure 12.4: Mouth ulcer following DMPS treatment (see Chapter 54).
EXAMPLES OF ALLERGIC AND OTHER ADVERSE DRUG REACTIONS 69predisposes to non-immune haemolysis (e.g. primaquine).Immune mechanisms include the following:1. Combination of the drug with the red-cell membrane,with the conjugate acting as an antigen. This has beenshown to occur with penicillin-induced haemolysis, andmay also occur with chlorpromazine and sulphonamides.2. Alteration of the red-cell membrane by the drug so that itbecomes autoimmunogenic. This may happen withmethyldopa, and a direct positive Coombs’ test developsin about 20% of patients who have been treated with thisdrug for more than one year. Frank haemolysis occurs inonly a small proportion of cases. Similar changes can takeplace with levodopa, mefenamic acid and beta-lactamantibiotics.3. Non-specific binding of plasma protein to red cells, andthus causing haemolysis. This is believed to occur withcephalosporins.Aplastic anaemia as an isolated entity is not common, butmay occur either in isolation or as part of a general depressionof bone marrow activity (pancytopenia). Examples includechloramphenicol and (commonly and predictably) cytotoxicdrugs.Agranulocytosis can be caused by many drugs. Several different mechanisms are implicated, and it is not knownwhether allergy plays a part. The drugs most frequently impli-cated include the following:• most cytotoxic drugs (Chapter 48);• antithyroid drugs (methimazole, carbimazole,propylthiouracil; Chapter 38);• sulphonamides and sulphonylureas (e.g. tolbutamide,glipizide; Chapter 37);• antidepressants (especially mianserin; Chapter 20) andantipsychotics (e.g. phenothiazines, clozapine; Chapter 20);• anti-epileptic drugs (e.g. carbamazepine, felbamate;Chapter 22).SYSTEMIC LUPUS ERYTHEMATOSUSSeveral drugs (including procainamide, isoniazid, hydralazine,chlorpromazine and anticonvulsants) produce a syndromethat resembles systemic lupus together with a positive anti-nuclear factor test. The development of this is closely relatedto dose, and in the case of hydralazine it also depends on therate of acetylation, which is genetically controlled (Chapter14). There is some evidence that the drugs act as haptens, com-bining with DNA and forming antigens. Symptoms usuallydisappear when the drug is stopped, but recovery may be slow.VASCULITISBoth acute and chronic vasculitis can result from taking drugs, and may have an allergic basis. Acute vasculitis withpurpura and renal involvement occurs with penicillins,sulphonamides and penicillamine. A more chronic form canoccur with phenytoin.RENAL DYSFUNCTIONAll clinical manifestations of renal disease can be caused bydrugs, and common culprits are non-steroidal anti-inflammatorydrugs and angiotensin-converting enzyme inhibitors (whichcause functional and usually reversible renal failure in suscep-tible patients; Chapters 26 and 28). Nephrotic syndromeresults from several drugs (e.g. penicillamine, high-dose cap-topril, gold salts) which cause various immune-mediatedglomerular injuries. Interstitial nephritis can be caused by sev-eral drugs, including non-steroidal anti-inflammatory drugsand penicillins, especially meticillin. Cisplatin, aminoglyco-sides, amphotericin, radiocontrast media and vancomycincause direct tubular toxicity. Many drugs cause electrolyte oracid-base disturbances via their predictable direct or indirecteffects on renal electrolyte excretion (e.g. hypokalaemia andhypomagnesaemia from loop diuretics, hyperkalaemia frompotassium-sparing diuretics, converting enzyme inhibitorsand angiotensin II receptor antagonists, proximal renal tubular acidosis from carbonic anhydrase inhibitors), andsome cause unpredictable toxic effects on acid-base balance(e.g. distal renal tubular acidosis from amphotericin).Obstructive uropathy can be caused by uric acid crystals con-sequent upon initiation of chemotherapy in patients withhaematological malignancy, and – rarely – poorly solubledrugs, such as sulphonamides, methotrexate or indinavir, cancause crystalluria.OTHER REACTIONSFever is a common manifestation of drug allergy, and should be remembered in patients with fever of unknowncause.Liver damage (hepatitis with or without obstructive fea-tures) as a side effect of drugs is important. It may be insidi-ous, leading slowly to end-stage cirrhosis (e.g. during chronictreatment with methotrexate) or acute and fulminant (as insome cases of isoniazid, halothane or phenytoin hepatitis).Chlorpromazine or erythromycin may cause liver involve-ment characterized by raised alkaline phosphatase and biliru-bin (‘obstructive’ pattern). Gallstones (and mechanicalobstruction) can be caused by fibrates and other lipid-loweringdrugs (Chapter 27), and by octreotide, a somatostatin ana-logue used to treat a variety of enteropancreatic tumours,including carcinoid syndrome and VIPomas (vasoactive intes-tinal polypeptide) (see Chapter 42). Immune mechanisms areimplicated in some forms of hepatic injury by drugs, but areseldom solely responsible.
70 ADVERSE DRUG REACTIONSFURTHER READING AND WEB MATERIALDavies DM, Ferner RE de Glanville H. Textbook of adverse drug reac-tions, 5th edn. Oxford: Oxford Medical Publications, 1998.Dukes MNG, Aronson JA: 2000: Meylers’s side-effects of drugs, vol. 14.Amsterdam: Elsevier (see also companion volumes Side-effects ofdrugs annuals, 2003, published annually since 1977).FDA Medwatch website. www.fda.gov/medwatchGruchalla RS, Pirmohamed M. Antibiotic allergy. New England Journalof Medicine 2006; 354: 601–609 (practical clinical approach).Howard RL, Avery AJ, Slavenburg S et al. Which drugs cause prevent-able admissions to hospital? A systematic review. British Journal ofClinical Pharmacology 2006; 63: 136–47.MHRA and the Committee on Safety of Medicines and the MedicineControl Agency. Current problems in pharmacovigilance. London:Committee on Safety of Medicines and the Medicine ControlAgency. (Students are advised to monitor this publication forongoing and future adverse reactions.)MHRA Current problems in pharmacovigilance website.www.mhra.gov.uk/home/idcplg?IdcServiceSS_GET_PAGE&nodeId368.Pirmohamed M, James S, Meakin S et al. Adverse drug reactions ascause of admission to hospital: prospective analysis of 18.820patients. British Medical Journal 2004; 329: 15–19.Rawlins MD, Thompson JW. Pathogenesis of adverse drug reactions, 2nd edn. Oxford: Oxford University Press, 1977.Case historyA 73-year-old man develops severe shoulder pain and isdiagnosed as having a frozen shoulder, for which he is pre-scribed physiotherapy and given naproxen, 250 mg threetimes a day, by his family practitioner. The practitioner knowshim well and checks that he has normal renal function forhis age. When he attends for review about two weeks later,he is complaining of tiredness and reduced urine frequency.Over the past few days he noted painful but non-swollenjoints and a maculopapular rash on his trunk and limbs. Heis afebrile and apart from the rash there are no otherabnormal physical signs. Laboratory studies show a normalfull blood count; an absolute eosinophil count raised at490/mm3. His serum creatinine was 110 μmol/L at baselineand is now 350 μmol/L with a urea of 22.5 mmol/L; elec-trolytes and liver function tests are normal. Urinalysisshows 2 protein, urine microscopy contains 100 leuko-cytes/hpf with 24% eosinophils.Question 1If this is an adverse drug reaction, what type of reaction isit and what is the diagnosis?Question 2What is the best management plan and should this patientever receive naproxen again?Answer 1The patient has developed an acute interstitial nephritis,probably secondary to the recent introduction of naproxentreatment. This is a well-recognized syndrome, with theclinical features that the patient displays in this case. It canbe associated with many NSAIDs (both non selective NSAIDsand COX-2 inhibitors), particularly in the elderly. This is atype B adverse drug reaction whose pathophysiology isprobably a combination of type III and type IV hypersensi-tivity reactions.Answer 2Discontinuation of the offending agent is vital and this is sometimes sufficient to produce a return to baseline values of renal function and the disappearance of systemic symptoms of fever and the rash. Recovery may possibly beaccelerated and further renal toxicity minimized by a shortcourse (five to seven days) of high-dose oral corticosteroids,while monitoring renal function. The offending agentshould not be used again in this patient unless the benefitsof using it vastly outweigh the risks associated with its usein a serious illness.