Paracetamol and Phenacetin

Drugs 32 (Suppl. 4): 46-59 (1986) OOI2-6667/86/040

Stephen P. Clissold ADIS Drug Information Services, Auckland

Summary Since their synthesis in the late l800s paracetamol (acetaminophen) and phenacetin have followed divergent pathways with regard to their popularity as mild analgesic/anti pyretic drugs. Initially, paracetamol was discarded in favour of phenacetin because the latter drug was supposedly less toxic. Today the opposite is true. and paracetamol. along with aspirin. has become one ofthe two most popular 'over-the-counter' non-narcotic anal gesic agents. This marked increase in the wide approval attained by paracetamol has been accompanied by the virtual commercial demise of phenacetin because of its role, albeit somewhat circumstantial. in causing analgesic nephropathy. Both paracetamol and phenacetin are effective mild analgesics, suitable for treating mild to moderate pain. and their actions are broadly comparable with those of aspirin and related salicylates. although they do not appear to possess significant anti-inflam matory activity. Since a major portion of a dose of phenacetin is rapidly metabolised to paracetamol. it seems possible that phenacetin owes some of its therapeutic activity to its main metabolite. paracetamol. whereas its most troublesome side effect (methaemoglo binaemia) is due to another metabolite. p-phenetidine. The mechanism of action of para cetamol is poorly defined. although it has been speculated that it may selectively inhibit prostaglandin production in the central nervous system. which would account for its anal gesic/antipyretic properties. The lack of any significant influence on peripheral cyclo oxygenase would explain the absence of anti-inflammatory activity. At therapeutic doses paracetamol is well tolerated and produces fewer side effects than aspirin. The most frequently reported adverse effect associated with paracetamol is he patotoxicity. which occurs after acute overdosage (usually doses greater than 10 to 15g are needed) and. very rarely. during long term treatment with doses at the higher levels ofthe therapeutic range. Paracetamol damages the liver through the formation ofa highly reactive metabolite which is normally inactivated by conjugation with glutathione. Ov erdoses of paraceta mol exhaust glutathione stores. thus allowing the accumulation ofthis toxic metabolite which covalently binds with vital cell elements and can result in liver necrosis. Glutathione precursors (notably intravenous N-acetylcysteine) have proved re markably successful in treating paracetamol overdose, as long as treatment is initiated within 10 hours. Apart from pharmacokinetic drug interactions resulting from changes in gastric empty ing rate. there have been very few reports of clinically important drug-drug interactions involving paracetamol or phenacetin. The pre-eminent position of paracetamol and aspirin as the non-narcotic analgesic agents of choice for mild to moderate pain is being seriously challenged by 'newer' non steroidal anti-inflammatory drugs. Some of these newer agents have been found to have significantly superior analgesic activity and. in some cases. longer durations of action. Paracetamol and Phenacetin 47

Paracetamol (acetaminophen) and phenacetin persuasive arguments. However, the case for phen are both derivatives of acetanilide (for structural acetin is weakened by its apparent lack of any ther formulae see fig. I), an aniline-like compound apeutic advantages over paracetamol, the potential which, in the 1880s, was serendipitously found to haematological toxicity of its metabolites, and its possess antipyretic activity. Acetanilide was quickly propensity (unlike other non-narcotic analgesics) introduced into medical practice and it was shown to produce central psychotropic effects which may to have analgesic as well as antipyretic activity. contribute to its liability for abuse. However, it was soon found to have unacceptable toxic effects, the most alarming being cyanosis due 1. Pharmacodynamic Properties to methaemoglobinaemia, and this prompted the 1.1 Mechanism of Action search for less toxic derivatives of aniline. Phen acetin was introduced into clinical use in 1887 by The basic pharmacological mechanisms of ac von Mering (von Mering, 1893), who also evalu tion of paracetamol and phenacetin have not re ated the activity of a related compound N-acetyl ceived the scientific attention accorded to the sali p-aminophenol (acetaminophen, paracetamol). He cylates and, consequently, many explanations for discarded paracetamol in favour of phenacetin be their activity appear somewhat speculative. Both cause of the latter's supposedly better toxicity pro drugs have analgesic and antipyretic properties file (for an overview of the history and usage of which do not differ significantly from those of as- the aniline derivatives see Bowman and Rand, 1980; Flower et ai., 1980; Meredith and Goulding, 1980). Further studies in the 1940s demonstrated that paracetamol was the major metabolite of both phenacetin and acetanilide and, importantly, met haemoglobinaemia was caused by a different me Acetanilide tabolite of phenacetin, p-phenetidine (Brodie and Axelrod, 1949). Paracetamol has been freely avail able since the mid 1950s and it has steadily gained in popularity. In the United Kingdom paracetamol sales have exceeded those of aspirin since about 1978. Paracetamol has never been quite so popular in the United States, although this position is Paracetamol (acetaminophen) changing and it has recently been reported that one proprietary paracetamol product accounts for about 35% of the 'over-the-counter' analgesic market (see Black, 1984). The rise in popularity of paracetamol has been accompanied by the virtual commercial demise of phenacetin. The major reason for this has been the condemnation of phenacetin, on what now seems Phenacetin to be circumstantial evidence, as being the sole cause of analgesic nephropathy (see section 4.2.1) [see also Kincaid-Smith, this issue]. Protagonists for and against the decisions taken in many coun tries to limit the supply of phenacetin are abundant Fig. 1. Structural formulae of acetanilide, paracetamol and within scientific circles, and each 'side' has its own phenacetin. Paracetamol and Phenacetin 48

pirin. However, the 2 drugs lack the potent anti demonstrated that paracetamol relieved pain by inflammatory actions of aspirin. blocking impulse generation at bradykinin-sensi Why paracetamol is an effective analgesic but tive chemoreceptors which evoke pain - a peri only a weak anti-inflammatory agent has not been pheral mechanism. Clearly, further studies are nec satisfactorily established, although most recent ex essary to evaluate the relative influence of planations involve a selective inhibition of some peripheral and central mechanisms on the overall facet of prostaglandin biosynthesis (for reviews see analgesic properties of paracetamol and phenace Hower et aI., 1980; Jackson et aI., 1984; Meredith tin. and Goulding, 1980; Ramwell, 1981). Some evi At the present time the mechanism of action of dence suggests that paracetamol has a weak inhib paracetamol is poorly understood; it is possible that itory influence on peripheral prostaglandin biosyn unknown peripheral and/or central mechanisms thesis (which would account for its lack of may also be involved. substantial anti-inflammatory activity), but that it is a potent inhibitor of prostaglandin production 2. Pharmacokilletic Properties within the central nervous system (presumably ac counting for its analgesic and antipyretic proper Acetanilide, phenacetin and paracetamol are all ties). Why central nervous system cydo-oxygenase derivatives of aniline, in whose structure reside the is more sensitive to paracetamol (if this is its main intrinsic antipyretic and analgesic properties of this mode of action) than peripheral cydo-oxygenase, group. Paracetamol is the major metabolite of both has not been convincingly explained and further acetanilide and phenacetin (see section 2.3) and the research is needed to clarify the mechanisms in pharmacodynamic actions of this compound prob volved. The analgesic and antipyretic effects of ably account for much of the therapeutic action of phenacetin are generally attributed to its major these drugs. Consequently, most attention in this metabolite, paracetamol; however, there is evi section will be focused on the pharmacokinetic dence that phenacetin itself has inherent pharma characteristics of paracetamol, a moderately water codynamic activity. and lipid-soluble weak organic acid (pKa 9.5) which is largely un-ionised over the physiological range 1.2 Analgesic Effects of pH.

Like the salicylates, paracetamol and phenace 2.1 Absorption tin possess analgesic activity which is effective against pain of mild to moderate severity. How Paracetamol is invariably taken orally and is ever, unlike the salicylates, which act mainly peri only minimally absorbed from the stomach. Ab pherally against pain associated with inflammation sorption is by passive diffusion with first-order ki (see Oissold, this issue, p. 8), both paracetamol and netics and occurs mainly in the small intestine; the phenacetin have few or no antiinflammatory prop rate of absorption therefore depends on the gastric erties and apparently exert their analgesic effects emptying rate (for reviews see Forrest et aI., 1982; via central actions (Bowman and Rand, 1980; Prescott, 1980). Hower et aI., 1980; Jackson et aI., 1984). This is Over all, paracetamol is rapidly and completely supported by the work of Ferreira et aL (1978) [see absorbed (fig. 2), although systemic bioavailability Ferreira, 1981] which suggests that release of pros after oral administration is incomplete owing to taglandins within the central nervous system, in first-pass metabolism. The proportion of a dose volving central pain circuits, as well as sensitisa reaching the systemic circulation as unchanged tion of peripheral pain receptors by locally released paracetamol appears to depend on the amount prostaglandins both contribute to inflammatory given, decreasing from about 90% after 1 to 2g to hyperalgesia. However, Guzman and Lim (1967) about 70% after 0.5g (Rawlins et aI., 1977). In fast- Paracetamol and Phenacetin 49

2.2 Distribution

50 Paracetamol is rapidly and uniformly distrib uted throughout body tissues; it achieves a tis sue: plasma concentration ratio of unity in most _ 20 ...J tissues except fat and cerebrospinal fluid. Follow Cl §.-- ing intravenous administration of paracetamol 12 c: mg/kg, the plasma concentration-time curve is bi 0 10 exponential (consistent with an open 2-compart ~c: II) ment model) with a short distribution half-life (t,;,,, c:0 0 5 ranging between 3 and 19 minutes), which is in 0 "0 dicative of rapid tissue distribution (fig. 2). In the E same group of healthy subjects, paracetamol 12 mg/ ~0 E kg administered as an oral solution was quickly ab aI 2 C. sorbed (mean T max"" 15 min) and, after a short dis aI E tribution phase, plasma concentrations declined in aI'" ii: parallel with those after intravenous injection I I I I o 2 3 4 5 6 (Prescott, 1980) [fig. 2]. Generally, the apparent Time (hours) volume of distribution of paracetamol is about 1 L/kg and is similar in healthy subjects, the elderly, and various patient groups, including anephric Fig. 2. Mean plasma paracetamol concentration following intra patients. At plasma concentrations of less than 60 venous (e) and oral (0) administration of paracetamol 12 mgl mg/L, paracetamol does not apparently bind to kg to 4 healthy subjects on separate occasions (after Prescott, plasma proteins; at 90 mg/L protein binding was 1980, with permission). less than 5%; and after toxic doses (plasma par acetamol concentrations of up to 280 mgfL) pro tein binding varied from 8 to 43% with no corre ing, healthy subjects, peak plasma paracetamol lation between binding and plasma paracetamol concentrations occur within 15 to 30 minutes of concentration (Gazzard et aI., 1973). Evidently, be administration of an oral solution. Absorption from tween 10 and 20% of administered paracetamol is tablets is slower (mean peak plasma concentrations bound to red blood cells. usually occur after about 60 minutes) and there may Phenacetin undergoes extensive first-pass me be as much as an 80-fold variation in individual tabolism by small intestinal mucosa and the liver, plasma paracetamol concentrations measured 60 and up to 80% is converted to paracetamol. The minutes after administration of a therapeutic dose apparent volume of distribution of phenacetin has (Prescott, 1974). After 8 hours only small amounts been reported to range between 1 and 2.1 L/kg; of unchanged paracetamol are detectable in plasma. plasma protein binding is about 33% (Heel and Following oral administration, phenacetin ab Avery, 1980). sorption is greatly influenced by its formulation - especially particle size. Generally, peak plasma 2.3 Elimination phenacetin concentrations are recorded within 1 to 2 hours (Prescott et aI., 1968). Between 60 and 80% Only 2 to 5% of a therapeutic dose of paracet of a therapeutic dose of phenacetin is de-ethylated amol is excreted unchanged in the urine; the re to paracetamol, and peak concentrations of active mainder is predominantly metabolised by the liver. metabolite in plasma are usually attained after ap At therapeutic dosages, paracetamol is mainly me proximately 3 hours. tabolised (> 80%) to the glucuronide and sulphate ~

o o I II II '" HN-C- CH3 NH2 6- hydrolysis I • ¢tCH' 6~ o =-.r Acetanilide OCeHgOe ::s Glucuronic acid conjugate ~Q:i ,.o~ o ~~. II ~o" o ~C-CH3 II

Ys0 3H 6:~c~ Sulphuric acid Postulated toxic intermediate conjugate o o d;~ H II Approximately 8% N-C-CH 3 HN-C-CH3 of a dose Paracetamol glutathione excreted • ~ --- •• as cysteine and glutathione mercapturic acid oo OH conjugates in the o N-acetyl-p-benzoquinoneimine urine II HN-C-CH 3 to hydrOlYSiS. <3 Other minor metabolites of phenacetin include: OC2H5 OC2H5 2-hydroxyphenetidine, 2- and 3-hydroxyphenacetin and N-hydroxyphenacetin. Less than 0.5% of Phenacetin p-Phenetidine a dose is recovered in the urine as unchanged phenacetin

Fig. 3. Schematic representation of the metabolic pathways postulated to account for the elimination of therapeutic doses of acetanilide, paracetamol and phenacetin V> (after Bowman and Rand, 1980; Brune and Lanz, 1984; Hinson, 1983; Mitchell et al., 1974; Prescott, 1980). For further details see text, section 2.3. o Paracetamol and Phenacetin 51

conjugates (fig. 3). A lesser fraction (about 10%) is of urinary pH but appears to be weakly correlated converted by cytochrome P-450-dependent hepatic with urine flow rate. The highly polar sulphate and mixed-function oxidase to a highly reactive meta glucuronide conjugates of paracetamol are appar bolite (postulated to be N-acetyl-p-benzoquino ently actively secreted by the tubules, as indicated neimine). In turn, this metabolite is rapidly inac by their respective renal clearance rates of approx tivated by conjugation with reduced glutathione and imately 170 and 130 ml/min, and there is no cor excreted in the urine, after further metabolism, as relation with urine flow or pH. The renal clearance cysteine and mercapturic acid conjugates. Large of the sulphate conjugate of paracetamol is con overdoses of paracetamol can exhaust stores of glu centration dependent (Morris and Levy, 1984). tathione and leave the highly reactive intermediate Of a dose of phenacetin, 60 to 80% is rapidly to bind covalently with vital cell elements, which metabolised to paracetamol and follows the elim may result in acute hepatic necrosis. ination pathways described above. Minor meta Paracetamol metabolism is age and dose de bolic pathways include deacetylation and hydrox pendent; the values quoted in figure 3 refer to the ylation to form p-phenetidine, 2-hydroxy administration of 20 mg/kg to healthy young sub phenetidine, 2- and 3-hydroxyphenacetin and N jects. 85 to 95% of a therapeutic dose is recovered hydroxyphenacetin. In total, over a dozen metab in the urine within 24 hours, the plasma half-life olites of phenacetin have been isolated (Margetts, is about 2 (range 1.5 to 3.0) hours, and total body 1976). Less than 0.5% of a dose of phenacetin is clearance is approximately 5 ml/min/kg. In young recovered unchanged in the urine. The mean children and neonates glucuronide conjugation is plasma half-life of phenacetin is approximately I deficient and sulphate conjugation is the dominant hour. Phenacetin de-ethylation to its major metab pathway; additionally, plasma half-life may be olite paracetamol is dose dependent and geneti slightly prolonged (Prescott, 1980). The plasma half cally determined; in individuals with limited meta life may also be increased if large or toxic doses of bolic capacity a greater proportion of a phenacetin paracetamol are taken and in patients with severe dose may be converted to toxic metabolites (Flower liver dysfunction. However, plasma paracetamol et aI., 1980). concentrations and half-life are within the normal range in patients with chronic renal disease, al 3. Analgesic Efficacy though the sulphate and glucuronide conjugates ac cumulate because of reduced renal excretion. Like aspirin and related salicylates (see Clissold, After ingestion of toxic doses of paracetamol, this issue, p. 8), paracetamol and phenacetin (at metabolism is usually impaired and plasma half doses of 650 to 1300mg) are effective in relieving life prolonged. Sulphate conjugation seems to be mild to moderate pain such as headache, tooth saturated very quickly, even in the therapeutic dose ache, dysmenorrhoea, a variety of postsurgical range, and a greater proportion of the administered pains, postepisiotomy pain, etc. For severe pain, dose is then excreted in the urine as glucuronide stronger analgesics (usually centrally acting agents) conjugate (increased from about 55% to between are usually required. Unlike the salicylates and 60 and 75%). However, in poisoned patients with newer non-steroidal non-narcotic analgesics, par severe liver damage, the excretion of mercapturic acetamol and phenacetin are relatively ineffective acid and cysteine conjugates of paracetamol in against chronic pain associated with inflammatory creases, in part at the expense of sulphate for arthritic diseases. This is not surprising, since such mation. pain relief has been shown to be dependent upon As a moderately lipid-soluble weak organic acid, the anti-inflammatory properties of the non-nar paracetamol undergoes considerable glomerular fil cotic analgesics, and paracetamol and phenacetin tration with subsequent extensive tubular reab have little or no anti-inflammatory activity (see sorption. Excretion of paracetamol is independent sections l.l and 1.2). Paracetamol and Phenacetin 52

The data assessing the analgesic efficacy of par not arthritic pain - see sections l.l and 1.2) and acetamol are far less extensive than those for as have similar dose-response and time-effect curves pirin. Indeed, the vast majority of controlled clinical (Cooper, 1981, 1983; Mehlisch, 1983). Thus, par studies have compared paracetamol with salicyl acetamol would seem to be a suitable alternative ates in a variety of painful conditions - but mainly mild to moderate analgesic in patients in whom pain of a self-limiting nature such as postsurgical salicylates are contraindicated (e.g. those with gas pain (especially after removal of impacted molars), trointestinal disease, advanced liver disease, aspi episiotomy pain, and dental pain. Many studies rin hypersensitivity and prolongation of bleeding have included a placebo comparison, and paracet time). Indeed, the superior side effect profile of amol has generally proved to be a superior anal paracetamol suggests that it may be the analgesic gesic to placebo (for reviews see Cooper, 1981, of choice, particularly in children, for those painful 1983). conditions not associated with inflammatory dis ease. 3.1 Paracetamol vs Phenacetin 3.3 Comparisons with Other Non-Steroidal There are very few clinical studies in which the Anti-Inflammatory Drugs (NSAIDs) analgesic efficacy of phenacetin (alone) has been determined. The analgesic effect of phenacetin was Cooper (1983) has extensively reviewed clinical shown to be similar to that of aspirin and par studies (usually single-dose trials in acute pain con acetamol in postpartum patients (Lasagna et aI., ditions such as that associated with oral surgery) 1967). Moertel et ai. (1972) evaluated the analgesic evaluating the analgesic efficacy of aspirin, par efficacy of paracetamol, phenacetin and other mild acetamol and many 'new' NSAIDs. Over all, the analgesics in 57 patients with pain caused by an results clearly demonstrated that newer non-ster unresectable malignancy. In these patients par oidal drugs such as ibuprofen, indoprofen, zome acetamol and phenacetin were equianalgesic and pirac, diflunisal, etc. (see Brogden, this issue, p. 27) both drugs were significantly superior to placebo. [fig. 4] produced significantly greater analgesia than Phenacetin seems to have central psychotropic activity (manifest as CNS-depressant effects such as relaxation, drowsiness, euphoria, etc.) which may contribute to its liability for abuse (Eade and Las 1.5 agna, 1967; Prescott et aI., 1970). Such properties do not appear to contribute to the analgesic activ ity of phenacetin, but may explain why phenacetin containing analgesic combinations have figured prominently in reports of analgesic abuse. With this in mind, and because paracetamol seems to pro duce a lower incidence of adverse effects than does phenacetin, paracetamol appears a better choice when a mild to moderate analgesic is required.

3.2 Paracetamol vs Aspirin 2 4 6 8 10 12 Time (hours) Aspirin and paracetamol have been frequently compared in a variety of painful conditions and Fig. 4. Time-effect curves for paracetamol (-; n = 34). diflunisal 500mg (000; n = 32). diflunisal 1000mg (---; n = 32) and pla the general opinion is that the 2 drugs are equian cebo (..... ; n = 30) in a double-blind study in patients with algesic and equipotent in most types of pain (but pain following oral surgery (after Forbes et al.. 1982). Paracetamol and Phenacetin 53

paracetamol and aspirin. As a group these drugs 4.1 Gastrointestinal Effects offer other advantages in that some ofthem appear to have a faster onset of action (e.g. zomepirac) and Unlike the effects of aspirin, no assQciation has others have a considerably longer duration of effect been established between paracetamol administra (e.g. diflunisal and piroxicam) [fig. 4] compared tion and gastrointestinal erosions or occult loss of with paracetamol. blood (lvey, 1983; Jick, 1981; Meredith and Goulding, 1980). Indeed, Ivey (1983) suggests that paracetamol is the analgesic/antipyretic of choice 3.4 Comparisons with Other Analgesics for adults or children with a history of upper gas trointestinal disturbances, portal hypertension and In the study of Moertel et al. (1972), paraceta oesophageal varices, reflux oesophagitis, or a mol 650mg and phenacetin 650mg were equian bleeding tendency (see also Ivey, this issue). algesic with pentazocine 50mg, mefenamic acid 250mg and codeine 65mg in 57 patients with pain 4.2 Renal Effects caused by cancer. Comparisons between dipyrone and paracetamol have produced varying results de 4.2.1 Analgesic Nephropathy pending on the type of pain treated. Over all, the Phenacetin has been removed from many of the 2 drugs seem to be of similar analgesic efficacy world's markets because it was singled out as the (Mehlisch, 1983), although the duration of action main cause of analgesic nephropathy (sometimes of dipyrone is longer (Mukherjee and Sood, 1980). termed 'phenacetin nephropathy'). The scientific basis supporting the supposed role of phenacetin as the major aetiological agent in analgesic nephro 4. Side Effects pathy seems somewhat circumstantial, if not spu rious (for an extensive review see Prescott, 1982). At therapeutic dosages (0.5 to l.Og orally 3 or 4 As noted by Kincaid-Smith (this issue) the case times daily) both paracetamol and phenacetin are against other non-narcotic analgesics such as the usually well tolerated and no common adverse ef acidic anti-inflammatory drugs (e.g. aspirin and re fects have been established. lated salicylates) in the development of renal pap Because paracetamol and aspirin are the appar illary necrosis is more convincing than the evi ent 'giants' of the 'over-the-counter' mild anal dence against phenacetin. gesic/antipyretic market, it is not surprising that The literature evaluating individual drugs and they have been subjected to extensive comparison their role, if any, in analgesic nephropathy has be (for example, a special issue of Archives ofInternal come voluminous, and yet the subject remains Medicine 141, 1981 was entirely devoted to com contentious. What is clear is that analgesic abuse, paring these 2 drugs, with a major emphasis on usually involving analgesic combinations, is a potential deleterious effects). Over all, paracetamol causative factor in the development of chronic renal seems to cause fewer side effects than aspirin: it disease. Early results implicated phenacetin as a does not produce the incidence of gastrointestinal common ingredient in the analgesic combinations disturbances that aspirin does, it rarely produces consumed, and it was removed from non-prescrip hypersensitivity reactions, and it has not been im tion analgesic products in some countries. This ac plicated in Reye's syndrome. Both aspirin and tion has not been followed by the expected fall in paracetamol can potentially produce severe toxic mortality from analgesic nephropathy. There now reactions after acute overdosage and sometimes seems to be sufficient doubt surrounding the pre during long term treatment employing doses at the sumed primary role of phenacetin as a nephrotoxic higher levels of the therapeutic range (Addy, 1983; agent (Kincaid-Smith, 1970, 1978; Prescott, 1982) Hollister, 1981). that we should be concerned about the widespread Paracetamol and Phenacetin 54

availability of other non-narcotic analgesics, for ex belled as the prime aetiological agent for inducing ample, those which contain acidic NSAIDs such as uroepithelial tumours because of its long-standing the salicylates and particularly those which are connection with analgesic nephropathy. The role of combination products. other analgesics in the pathogenesis of urinary tract Paracetamol is the major metabolite of phena tumour formation has not been adequately as cetin and there has been considerable concern about sessed and there seems little justification, at pre its role in analgesic nephropathy. Over all, par sent, in ascribing primary responsibility to phen acetamol seems to have one of the lowest pro acetin. pensities to be nephrotoxic, although there have been case reports implicating it as the sole cause 4.3 Hepatic Effects of renal papillary necrosis (Segasothy et aI., 1984). In view of the discord regarding which drugs are When taken in large overdoses paracetamol and primarily responsible in the pathogenesis of anal phenacetin are hepatotoxic and this now appears gesic nephropathy, it is not surprising that the to be due to a toxic reactive intermediary metab mechanisms involved are still unknown. Inhibi olite (probably N-acetyl-p-benzoquinoneimine; see tion of cyclo-oxygenase and hence prostaglandin fig. 3) which accumulates as glutathione stores are production within the kidney has been postulated, depleted (see section 4.7). although a direct cytotoxic action of analgesics, or In healthy adults normal doses of paracetamol more likely intermediary metabolites, is also pos (up to 4 g/day), particularly when used occasion sible. Phenacetin and paracetamol (and the sali ally or over short periods, do not produce any ad cylates) are all converted in the kidney to poten verse hepatic effects. However, a few reports have tially cytotoxic metabolites but their relevance to raised the possibility that toxic hepatitis may ac the genesis of analgesic nephropathy is uncertain. company prolonged periods of paracetamol treat ment (apparently with therapeutic dosages) in 4.2.2 Urinary Tract Tumours healthy subjects, but more especially in patients Uroepithe1ial tumours of the renal pelvis asso with a history of alcohol abuse, previous hepatic ciated with renal papillary necrosis and analgesic disease or malnutrition, or who are taking drugs abuse were first reported by Hultengren et a1. (1965) that induce liver microsomal enzymes (Black, 1984; and represent a further serious long term compli Meredith and Goulding, 1980; Zimmerman, 1981). cation accompanying the overuse of non-narcotic Thus, in reviewing the literature Black (1984) re analgesics. The role of phenacetin in the develop ported data from 13 alcoholic patients who devel ment of tumours of the urinary tract is not clear, oped a hepatotoxic reaction to doses of paraceta though much has been made of the possible con mol from as little as 2.6g to greater than 109. Five version of phenacetin to potentially carcinogenic of the patients died and liver biopsy performed in metabolites (Prescott, 1982). 8 of the group revealed extensive centrilobular ne Generally, uroepithelial tumours have occurred crosis. It seems likely that the dosages admini in patients taking combinations of phenacetin with stered were excessive in patients who could rea aspirin or pyrazolones (Prescott, 1982); tumours sonably be expected to be 'at risk', and it is notable have also occurred, albeit rarely, in patients re that all the deaths occurred in patients taking 109 ceiving antipyrine and salicylates without phena or more of paracetamo1. Whether it was taken for cetin (see Kincaid-Smith, this issue). Such carci therapeutic purposes or not, it is likely that these nomas seem to develop in response to irritation or reactions represent acute toxic sequelae resulting injury from a large number of causes, and although from overdosage (whether intentional or other phenacetin may yet prove to be one of these causes, wise) in patients with enhanced susceptibility. available data do not convincingly support such an In a survey of 45 patients with chronic active association. Phenacetin seems to have been la- hepatitis, Neuberger and colleagues (1980) were Paracetamol and Phenacetin 55

unable to find any evidence to implicate paracet developing aspirin hypersensitivity (see p. 8 table amol as the cause of the disease. Additionally, III) paracetamol appears to be a relatively safe al paracetamol did not affect liver function tests in though not completely innocuous alternative. patients who periodically used the drug during fol Skin rash (erythematous or urticarial) occurs in low-up. frequently with paracetamol and phenacetin. Very From the above data it appears that if thera rarely, it is more serious and may be accompanied peutic dosages of paraceta mol can cause liver dam by mucosal lesions, laryngeal oedema and 'drug age, it must be an extremely rare event, and the fever'. mechanisms involved remain obscure. 4.6 Pregnancy and Lactation 4.4 Haematological Effects Available data do not indicate any specific con Paracetamol does not affect normal haemostatic traindications to the use of paracetamol during mechanisms and it is increasingly preferred to as pregnancy. No association was observed between pirin in patients with pre-existing bleeding disord paracetamol administration during pregnancy and ers. Very rarely paracetamol usage has been asso the resultant offspring with congenital abnormali ciated with various blood dyscrasias such as ties or stillbirths in a survey conducted by the Royal neutropenia, pancytopenia, leucopenia, thrombo College of General Practitioners (Crombie et at, cytopenia, methaemoglobinaemia, haemolytic an 1970). As with all drugs, paracetamol should be aemia and agranulocytosis (American Hospital used in pregnant women only if the expected bene Formulary, 1985; Flower et at, 1980). The num fits outweigh any potential risk. bers of patients involved have been extremely small Paracetamol is excreted in milk but only small - usually case reports. quantities are transferred. Consequently, usual In a few isolated cases phenacetin use has been therapeutic analgesic doses of paracetamol do not considered the cause of various blood dyscrasias seem to present a risk to nursing infants (Austra including aplastic anaemia and agranulocytosis. lian National Drug Information Service, 1984). Methaemoglobinaemia and haemolytic anaemia are relatively common complications associated with 4.7 Overdosage chronic ingestion of high doses of phenacetin but they rarely produce significant clinical conse Clinical symptoms of acute paracetamol over quences. Both are caused by toxic metabolites of dose are divisible into 3 phases, although less sev phenacetin; p-phenetidine has long been consid ere cases may not progress through all 3 (Austra ered responsible although it may only be a precur lian National Drug Information Service, 1984; sor for other oxidative aromatic amine metabolites Prescott, 1983; Rumack, 1983). During the first 24 (Flower et at, 1980). hours (phase 1) there are no specific early symp toms and patients may not seek medical attention. 4.5 Hypersensitivity However, anorexia, nausea, vomiting and malaise may develop and persist. Between 24 and 48 hours Paracetamol is far less likely to produce a after ingestion of the overdose (phase 2), the symp bronchospastic or urticarial hypersensitivity reac toms evident in phase 1 usually become less severe tion than aspirin and it has been recommended as and right upper quadrant and abdominal pain may a suitable substitute for the salicylate in susceptible occur. Liver function tests become abnormal at this individuals. However, a small proportion of stage, but maximum derangement does not occur patients (5 to 6%) who are sensitive to aspirin ex until at least the third day. Urine output may de hibit a cross-sensitivity to paracetamol (Settipane, crease owing to dehydration, the antidiuretic effect 1981). Thus, for those patients at particular risk of of paracetamol or, rarely, renal damage. Paracetamol and Phenacetin 56

Not interpretable Treatment Treat if above lower line under 4h useless

300 , , " , , 200 , , Severe liver , ,~amage , likely Mild' , ", 100 , , E " , ~ c:: " , o " , ~tl 50 c:: " 8 " "0 E as fi ~ 30 ~ as .,E as ~ 20~------+------'------r------+------~ o 4 8 12 15 Time aiter ingestion of overdose (hours)

Fig. 5. Relationship between plasma concentrations. time aiter ingestion. and liver damage following paracetamol overdosage. Spe cific antidotal therapy is indicated above the solid 'treatment' line (aiter Prescott, 1983).

Phase 3 occurs, if at all, after 3 to 5 days as a Without specific treatment, severe liver damage consequence of extensive acute hepatic necrosis. following paracetamol overdose occurs in only ap Liver failure with dramatic increases in plasma as proximately 8% of patients; about 1% develop renal partate (AST) and alanine (ALT) aminotransferase failure and between 1 and 2% die from hepatic fail activity (to 10,000 U/L or more), with or without ure (Prescott, 1983). Since the extent of poisoning jaundice, hypoglycaemia and coagulation defects, cannot reliably be assessed by clinical or biochem may ensue. In less severely affected patients liver ical methods during the early phases following necrosis is localised to centrilobular areas but in paracetamol overdose, plasma paracetamol con severe disease, with fulminant hepatic failure, wide centrations in relation to time after ingestion have areas of confluent necrosis develop. Fulminating been used to assess potential risk. Prescott (1983) liver failure with deepening jaundice and coma suggests that a line joining a plasma paracetamol carries a poor prognosis but occurs in only a small concentration of 200 mgfL at 4 hours and 30 mgf minority of patients. If death does not occur, nor L at 15 hours can be used as a guide to predict the mal liver function returns over a few weeks. severity of liver damage (fig. 5). Paracetamol and Phenacetin 57

Paracetamol produces liver damage through the tric emptying rate, there have been very few re accumulation of a highly reactive metabolite ports of clinically important drug-drug interactions (probably N-acetyl-p-benzoquinoneimine) which is involving paracetamol or phenacetin (Hull Hayes, formed by cytochrome P-4S0-dependent mixed 1981; Prescott, 1980). function oxidase (Jollow et ai., 1974; Mitchell et Paracetamol is sometimes used as a model for ai., 1973, 1974). This arylating metabolite is nor drug absorption studies because it is a weak acid mally inactivated by preferential conjugation with that is largely un-ionised in both gastric and intes hepatic glutathione and excreted as cysteine and tinal fluid and its rate of absorption is directly re mercapturic acid conjugates after further metab lated to the gastric emptying rate. Drugs which af olism. It has been estimated that 10 to ISg of par fect gastric emptying rate influence the rate, but acetamol is needed to substantially deplete liver not the extent, of paracetamol absorption. Sub glutathione stores and so threaten hepatic function, stances shown to delay gastric emptying rate and, although, occasionally, lower doses have been im consequently, slow the rate of paracetamol absorp plicated (Prescott, 1983). tion include drugs such as propantheline (an anti The above basic description of the underlying cholinergic compound used in the treatment of pathogenesis involved in paracetamol-induced he peptic ulcers) and narcotic analgesics [pethidine patotoxicity has been used to develop a suitable (meperidine), heroin (diamorphine) and pentazo means of treatment. Of the methods used, admin cine], and certain foodstuffs (especially carbohy istration of precursors of glutathione have proved drates). Metoclopramide accelerates gastric empty remarkably successful in preventing liver damage. ing and increases the rate of paracetamol absorption Intravenous N-acetylcysteine seems to be the most (the peak plasma paracetamol concentration is also effective choice but must be given within 10 hours increased). of ingestion of the paracetamol overdose; no bene It has been suggested that drugs known to in fit is gained if treatment is delayed beyond 12 to duce hepatic microsomal enzymes (e.g. ethanol, IS hours (Prescott, 1983). phenobarbitone, phenytoin, meprobamate) may Adults account for the vast majority of serious reduce the margin of safety for paracetamol-in and fatal cases of paracetamol poisoning and there duced hepatotoxicity owing to an increased pro is some evidence that young children are resistant duction of hepatotoxic metabolites via the micro to its hepatotoxic effects (Peterson and Rumack, somal system (Australian National Drug 1981; Rumack, 1984). However, instances of sev Information Service, 1984). However, recent stud ere and fatal liver damage in children have been ies have shown that anticonvulsants, rifampicin and reported (Arena et aI., 1978; Greene et aI., 1983; ethanol do not increase the formation of hepato Nogen and Bremner, 1978). toxic metabolites of paracetamol in man (Prescott Although paracetamol is the major metabolite and Critchley, 1983). Isolated studies suggest par of phenacetin, the toxic manifestations with phen acetamol may lengthen prothrombin time in acetin overdosage are markedly different. Thus, patients taking anticoagulants. However, the ef phenacetin may cause methaemoglobinaemia, fects seem to be of little clinical consequence and haemolytic anaemia and renal tubular necrosis, and paracetamol appears to be preferable to salicylates lethal doses are not associated with liver necrosis in this setting when a mild analgesic or antipyretic but more usually with cyanosis, respiratory depres is needed. Paracetamol increases plasma chlor sion and cardiac arrest (Flower et aI., 1980). amphenicol half-life and a reduction in the dosage of antibiotic may be necessary if the 2 drugs have 5. Drug Interactions to be given concomitantly. A number of theoretical or speculated drug in Apart from pharmacokinetic drug interactions teractions (J3-blockers and chlorpromazine) are of resulting from a delay or a stimulation of the gas- dubious clinical relevance and await further inves- Paracetamol and Phenacetin 58

tigation. It could be clinically prudent to avoid high Greene, J.W.: Craft, L. and Ghishan, F.K.: Acetaminophen poi doses of paracetamol in patients taking drugs with soning in infancy. American Journal of Diseases in Children 137: 386-387 (1983). similar toxicities (hepatotoxicity, haemolytic an Guzman, F. and Lim, R.K.S.: Central and peripheral mechanisms aemia, methaemoglobinaemia and perhaps neph of analgesia and pain. Archivos de Biologica y Medicina Ex perimentalis 4: 180-186 (1967). rotoxicity), although there does not seem to be any Heel, R.C and Avery, G.S.: Drug data information; in Avery evidence that an exacerbation of toxic effects oc (Ed.) Drug Treatment (2nd ed.) pp. 1211-1222 (ADIS Press, Auckland 1980). curs when paracetamol is used with such drugs. Hinson, J.A.: Reactive metabolites of phenacetin and acetami nophen. Environmental Health Perspectives 49: 71-79 (1983). Hollister, L.E.: Perspectives and summation of symposium, as pirin and acetaminophen. Archives of Internal Medicine 141: References 404-406 (1981). Hull Hayes, A.: Therapeutic implications of drug interactions with Addy, D.P.: Cold comfort for hot children. British Medical Jour acetaminophen and aspirin. Archives of Internal Medicine 141: nal 286: 1163-1164 (1983). 301-304 (1981). American Hospital Formulary: Miscellaneous analgesics and an Hultengren, N.; Lagergren, C and Ljungqvist, A.: Carcinoma of tipyretics; in Drug Information 85, pp. 802-806 (American So the renal pelvis in renal papillary necrosis. Acta Chirurgica ciety of Hospital Pharmacists, Bethesda 1985). Scandinavica 130: 314-320 (1965). Arena, J.M.; Rourk, M.H. and Sibrack, CD.: Acetaminophen: Ivey, K.I.: Gastrointestinal effects of antipyretic analgesics. report of unusual poisoning. Pediatrics 61: 68-72 (1978). American Journal of Medicine 75: 53-64 (1983). Australian National Drug Information Service: Profile on para Jackson, CH.: MacDonald, N.C and Cornett, J.W.D.: Aceta cetamol. Australian Journal of Pharmacy 776: 857-863 (1984). minophen. Canadian Medical Association Journal 131: 25-37 Black, M.: Acetaminophen hepatotoxicity. Annual Reviews of (1984). Medicine 35: 577-593 (1984). Jick, H.: Effects of aspirin and acetaminophen in gastrointestinal Bowman, W.C and Rand, M.I. (Eds): Textbook of Pharmacology hemorrhage. Archives ofinternal Medicine 141: 316-321 (1981). (2nd ed.) [Blackwell Scientific Publications, Oxford 1980]. Jollow, DJ.; Thorgeirsson, S.S.; Potter, W.Z.; Hashimoto, M. and Brodie, B.B. and Axelrod, J.: The fate of acetophenetidin (phen Mitchell, J .R.: Acetaminophen-induced hepatic necrosis. VI. acetin) in man and methods for the estimation of acetophe Metabolic disposition of toxic and nontoxic doses of aceta netidin and its metabolites in biological material. Journal of minophen. Pharmacology 12: 251-271 (1974). Pharmacology and Experimental Therapeutics 97: 58-67 (1949). Kincaid-Smith, P.: Analgesic nephropathy. British Medical Jour Brune, K. and Lanz, R.: Nonopioid analgesics; in Kurhar and nal 282: 1790-1791 (1970). Pasturnak (Eds) Analgesics: Neurochemical Behavioral and Kincaid-Smith, P.: Drug-induced renal disease. Practitioner 220: Clinical Perspectives (Raven Press, New York 1984). 862-867 (1978). Cooper, S.A.: Comparative analgesic efficacies of aspirin and Lasagna, L.; Davis, M. and Pearson, J.W.: A comparison of ace acetaminophen. Archives of Internal Medicine 141: 282-285 tophenetidin and acetaminophen I. Analgesic effects in post (1981). partum patients. Journal of Pharmacology and Experimental Cooper, S.A.: New peripherally-acting oral analgesic agent. An Therapeutics 155: 296-300 (1967). nual Review of Pharmacology and Toxicology 23: 617-647 Margetts, G.: Phenacetin and paracetamol. Journal of Interna ( 1983). tional Medical Research 4: 55-70 (1976). Crombie, D.L.; Pinsent, R.I.F.H.; Slater, B.C; Reming, D. and Mehlisch, D.R.: Review of the comparative analgesic efficacy of Cross, K.W.: Teratogenic drugs - R.CG.P. survey. British salicylates, acetaminophen, and pyrazolones. American Jour Medical Journal 4: 178-179 (1970). nal of Medicine 75: 47-52 (1983). Eade, N.R. and Lasagna, L.: A comparison of acetophenetidin Meredith, T.J. and Goulding, R.: Paracetamol. Postgraduate and acetaminophen II. Subjective effects in healthy volunteers. Medical Journal 56: 459-473 (1980). Journal of Pharmacology and Experimental Therapeutics 155: Mitchell, J.R.; Jollow, DJ.; Potter, W.Z.; Davis, D.C.; Gillette, 301-308 (1967). J.R. and Brodie, B.B.: Acetaminophen-induced hepatic necro Ferreira, S.H.: Peripheral and central analgesia. Pain I: 54 (1981). sis. IV. Role of drug metabolism. Journal of Pharmacology and Ferreira, S.H.: Lorenzetti, B.B. and Correa, F.M.A.: Central and Experimental Therapeutics 187: 185-194 (1973). peripheral anti-algesic action of aspirin-like drugs. European Mitchell, J.R.; Thorgiersson, S.S.: Potter, W.Z.; Jollow, DJ. and Journal of Pharmacology 53: 39-48 (1978). Keiser, H.: Acetaminophen-induced hepatic injury. Protective Flower, R.I.: Moncada, S. and Vane, J.R.: Analgesic-antipyretics role of glutathione in man and rationale for therapy. Clinical and anti-inflammatory agents; drugs employed in the treat Pharmacology and Therapeutics 16: 676-684 (1974). ment of gout: in Goodman and Gilmans (Eds) The Pharma Moertel, CG.; Ahmann, D.L.; Taylor, W.F. and Schwartau, N.: cological Basis of Therapeutics (6th ed.) pp. 682-728 (Mac A comparative evaluation of marketed analgesic drugs. New millan Publishing Company, New York 1980). England Journal of Medicine 286: 813-815 (1972). Forbes, J.A.: Beaver, W.T.; White, E.H.; White, R.W.; Neilson, Morris, M.E. and Levy, G.: Renal clearance and serum protein G.B. and Shackleford, R.W.: A new oral analgesic with an un binding of acetaminophen and its major conjugates in hu usually long duration of action. Journal of the American Medi mans. Journal of Pharmaceutical Sciences 73: 1038-1041 (1984). cal Association 248: 2139-2142 (1982). Mukherjee, S. and Sood. S.: A controlled evaluation of orally ad Forrest, J.A.H.; Clements, J.A. and Prescott, L.F.: Clinical phar ministered aspirin, dipyrone and placebo in patients with post macokinetics of paracetamol. Clinical Pharmacokinetics 7: 93- operative pain. Current Medical Research and Opinion 6: 619- 107 (1982). 623 (1980). Gazzard, B.G.: Ford-Hutchinson, A.W.; Smith, MJ.H. and Wil Neuberger. J.; Davis, M. and Williams, R.: Long-term ingestion liams, R.: The binding of paracetamol to plasma proteins of of paracetamol and liver disease. Journal of the Royal Society man and pig. Journal of Pharmacy and Pharmacology 25: 964- of Medicine 73: 701-707 (1980). 967 (1973). Nogen, A.G. and Bremner, J.E.: Fatal acetaminophen overdosage Paracetamol and Phenacetin 59

in a young child. Journal of Pediatrics 92: 832-833 (1978). Rawlins, M.D.: Henderson, D.B. and Hijab, A.R.: Pharmacoki Peterson. R.G. and Rumack, B.H.: Age as a variable in aceta netics of paracetamol (acetaminophen) after intravenous and minophen overdose. Archives of Internal Medicine 141: 390- oral administration. European Journal of Clinical Pharmacol 393 (1981). ogy II: 283-286 (1977). Prescott, L.F.: Gastrointestinal absorption of drugs. Medical Clinics Rumack, B.H.: Acetaminophen overdose. American Journal of of North America 58: 907-916 (1974). Medicine 15: 104-112 (1983). Prescott, L.F.: Kinetics and metabolism of paraceta mol and phen Rumack. B.H.: Acetaminophen overdose in young children. acetin. British Journal of Clinical Pharmacology 10: 2915-2985 American Journal of Diseases in Children 138: 428-433 (1984). (1980). Segasothy, M.; Kong Chiew Tong, B.; Kamel, A.; Murad, Z. and Prescott. L.F.: Analgesic nephropathy. Drugs 23: 75-149 (1982). Suleiman, A.B.: Analgesic nephropathy associated with para Prescott, L.F.: Paracetamol overdosage. Drugs 25: 290-314 (1983). cetamol. Australian and New Zealand Journal of Medicine 14: Prescott, L.F. and Critchley, J.A.J.H.: Drug interactions affecting 23-26 (1984). analgesic toxicity. American Journal of Medicine 75 (Suppl. Settipane, G.A.: Adverse reactions to aspirin and related drugs. SA): 113-116 (1983). Archives of Internal Medicine 141: 328-332 (1981). Prescott, L.F.; Sansur, M.; Levin, W. and Conney, A.H.: The Von Mering, J.: Beitriige zur Kenntniss der Antipyretica. Therap. comparative metabolism of phenacetin and N-acety-p-ami Monatsch 7: 577-587, (1893). nophenol in man, with particular reference to effects on the Zimmerman, H.J.: Effects of aspirin and acetaminophen on the kidney. Clinical Pharmacology and Therapeutics 9: 605-614 liver. Archives of Internal Medicine 141: 333-342 (1981). (1968). Prescott. L.F.: Steel, R.F. and Ferrier, W.R.: The effects of par ticle size on the absorption of phenacetin in man: a correlation between plasma concentration of phenacetin and effects on the central nervous system. Clinical Pharmacology and Therapeu tics II: 496-504 (1970). Ramwell, P.W.: Biologic importance of arachidonic acid. Ar Author's address: s.P. elisso/d, ADIS Drug Information Services, chives of Internal Medicine 141: 275-278 (1981). Private Bag, Mairangi Bay, Auckland 10 (New Zealand).