REVIEW ARTICLE EJP European Journal of Pain Acetaminophen/paracetamol: A history of errors, failures and false decisions K. Brune’, B. Renner’, G. Tiegs? 1 tesnute ot Expenmenta ara chaeal Pharmac 2 trate Conespondence ay Brute Ema breneapharmakologl unterangen ce Funding sources Contes of interests Nore dedared, Accepted for publication 22septenver 2014 ora r002%¢p.021 1. Discovery by error and serendipity 1.1 Acetanilide Experimental mmunelogy and Hepatology, Unversity Medal Cn and Toneatogy Une cerany ‘cemany Abstract Acclaminophen/paracetamol isthe most widely used drug of the world. At the same time, it is probably one of the most dangerous compounds in ‘medical use, causing hundreds of deaths in all industrialized countries due to acute liver fallure (ALF). Publications of the Tast 130 years found in the usual databases were analyzed, Personal contacts existed to renowned researchers having conitibuted to the medical use of paracetamol and its Precursors as H.U. Zollinger, S. Moeschilin, U, Dubaeh, 3. Axelrod an ethers. Further information is found in earlier reviews by Eichengran, Rodnan and Benedek, Sneader, Brune; comp. references. The history of the discovery of paracetamol staris with an error (active against worms), continues with a false assumption (paracetamol js saler than phenacetin), describes the first side-effect ‘epidemy’ (phenacetin nephropathy, drag induced interstitial nephritis) and ends with the discovery of secon generation problems due to the unavoidable production of a highly toxic metabolite of paracetamol N-acetyl-p-benzoquinone imine (NAPQI) that ‘may cause not only ALF and kidney damage butalso impaired development col the fetus and the newborn child, It appears timely to reassess the riskshenefit ratio of this campound dered why the antipyretic elect had not been described before, Researching the source of the ‘naph- thalene’, they learnt that (the local) pharmacy hhad provided them with acetanilide (Fig. 1) Instead In 1884, an important discovery took place in Stras- bourg (Germany, later France). At that time, the medical linic of Strasbourg was headed by the famous professor Adolf Kussmaul. Two young assistants, A. Cahn and P. Hepp, asked his advice paticnt who suflered from many ailments, among those fever and worm infestation (Cahn and Hepp, 1886). Kussmaull suggested to try naphthalene, which, had been recommended as an intestinal antiseptic (Sneadet, 1985). The two young physicians followed the advice and were disappointed by the result ~ most including the worms. Much to their surprise, they observed, however, that the lever of the patient fell shortly after the administration of the putative naphthalene. The two physicians won- how to treat a ailments remained, of naphthalene ~ an. error with long-lasting consequences! Consequently, Calin and Hepp tested acetanilide in rabbits and dogs (at that time, an animal ethics request was net needed) (Cahn and Hepp, 1886). They observed reliable but short-lasting aitipyretic activiiy: Both species did not sulfer from (visible, acute) toxie- ity. The young physicians then dared to try acetanilide in 24 patients (certainly without iniormed consent} ‘They were excited 9s all of them showed a reduction of fever. Acetanilide mumed om as elective as phenazone, which had been successtully put to the marketplace two years belore (Filelane, 1884). Most patients tolerated the treatment well, The only problem Cahn and Hepp observed was that some svar 19 a0) a52-965 958130 years of aniline analgesics What's already known about this topic? + Aniline-derived analgesics, particulatly paraceta mol, go along with a variety of unwanted drag effects. Fatal acute liver damage may occur at overdose and in at-isk groups. + Paracetamol interferes with the production of prostaglandins, This may further the develop- ‘ment of intestinal ulcers, kidney disfunetion and hypertension. Paracctamal is the only non-opioid analgesic rec- ‘ommended to pregnant women and their off- spring. Recent evidence links the use ol ‘Paracetamol during pregnaney and in newborns to an enhanced incidence of asthma, psycholoai- cal and physical development, and attention- Gelidisbyperactivity disorder (ADHD). + Experimental research demonstrates that immune regulation and the neuronal develop- ment during pregnancy may be distorted by paracetamol. ‘What does this review add? * This review shows that the discovery of the aniline analgesics was nol a consequence of sys- tematic research and cautious toxicological evaluation but rather the result of serendipity, certors and failures. The growing list of unwanted rug effects can be explained by the blockade of the production of prostaglandins and N-acetyl-p- benzoguinone imine (NAPQN). This highly reac- tive intermediate is produced in the mother and ‘most likely the fetus. Ie may be the culprit ofthe known hepatotoxicity and the cause of the new signals of second-genetation toxicity. + The production of NAPQI cannot be suppressed nor can its covalent binding to macromolecules be avoided. Consequently, paracetamol would not be admitted to the market today. To keep it as an over-the-counter (OTC) analgesic is ilogical and dangerous, patients developed a bluish color of the skin. This Phenomenon was not taken seriously. Enthusiasm Kranslated inte therapeutic action. A small company lose to Strasbourg, Kalle, produced large quantities of acetanilide for the young physicians. They distributed It to interested colleagues and soon ‘antlfebrin’, as acetanilide was now termed, was intensively used to reduce fever. Wider use went along with more reports of the bluish skin’, which was later identified to result from 954 carvan 190015 252- Brune etal methemoglobinaemia (Sneader, 2005: Prescott, 1996). Not surprisingly as the chemical aniline, irom which, acetanilide was synthesized, was alzeady known to be haematotoxic. It initiated methemoglobinaemla, thus Interfering with the oxygen transport by haemoglobi as many other aminophenols (Kiese and Pekis, 1964) Consequently, a search for safer chemical derivatives began. 1.2 Phenacetin 1 is to be credited 10 O, Hinsberg (1913) have synthesized phenacetin (Fig. 1). Together with a phar- macologist from Freiburg, A. Kast, he showed that the new compound worked as acctanilide but was less toxic (Hinsberg, 1913). In addition, it could be pro- duced at low costs, as Bayer and other companies were flooded with large amounts of para-nitraphenol Which is a by-product of the production of dianisidine (needed to produce the new synthetic “aniline dyes’) C. Duisberg (1913, director of Bayer) reported that Bayer was searching far customers of the useless para nitrophenol. He states that it was pure luck (Duisberg, 1913) that O. Hinsberg togetter with A. Kost found that phenacetin can be produced easfly from para nitrophenol and that iti both more effective and safer than acetanilide. ©, Hinsbeng and A. Eichengriin (the head chemist at Bayer), on the contrary, claimed that thorough thinking and the application of systematic synthetic chemistry led to the discovery of phenacetin (Eichengriin, 1913). In the end, it does not matter who contributed most to the discovery of phenacetin Duisberg or Hinsberg; in the end, it became a very successful drug (Fig, 2) Interestingly, two other new synthetic drugs, phenazone and aspirin, were also lound by serendipity and brought to the market under false assumptions. Aspirin, phenazone and phenacetin were synthesized in Germany and distsbuted quantities. worldwide in tons 1.3 Aspirin, phenazone and phenacetin, the first synthetic analgesics Aspirin (introduced 1899 by Bayer), phenazone (silehne, 1884) and phenacetin (Hinsberg, 1913) relieved pain and fever reliably, fast and without apparent dangers. They added (o the worldwide repa tatlon of the emerging German drug industry as described by Tainter (1948). He states: In 1884, Luitwig Kuors for « symtetiestbstiute for quinine, prepared antipytine, ie analgesic powers probably have sot been fully uilized while searching phenicone, a compound whose130 yeas ofan 1 analgesics Paso main] (C2) fseroncone | [ladder cancer? tvdcn, Rectaniide Metheroglobine ho Neo, Phenacetin ‘adoey damage Sccr, ° whos aden nde ( (coor) (2o-asx ‘SS “UDR lacurensar Piwnctutowenstease 7) / wnstersse dy Se \ ( dctueuronide ‘Acetaminophen 050, | * (Acetylaminophenol (APAP) Paracetamol) RENALERRETION] ——g.15y) | <—Crtoctwome [RENAL EXCRETION | ( Z pCa, (CrP2EA) ndn, tuathione [RPAP-mercaptate Figure 1 Avie arts meta sod madd ror Presat, 1903) Figure 2 The fathers of phenaceb (pictues 07. insber could no: be four. Sou lutathione-s: transferase Macetyl-p- bbenroquinoneimine(NAPQS) Yin, wpbos A Tf SProtein {felutathione << 30% Normal Hepatic Renal falure Developmental damage Covalent binding of amino acidsin proteinsané enzymes Duisberg Carl 1861-1935 Eichengrin Arthur 1867-1949 Von Mering Josef 1849-1908 955130 years of aniline analgesics Axelrod Julius 1912-2004 Brodie Bernard B. 1907-1989 Figure | Avckodand 8 Brodie investigated themessbolem of phen scat rd sugsered tht paracetarcl the safer dy Source. Wipe sia The ANINGO Bouman SF: Sog-aphy Seevon Knori’s synthesis of antipyrine mavked the beginning ofthe famous German drug industry anid ushered an Germany's forty-year dominance of the synttedic drug and chemical field, Aminopyrine (Pyranriden®), a sill more pote com por closely related to aritipyrine ced a few yours later in 1896 by Knorr and Stolz, This remained the most powerful analgesic of synthetic origin nit rece times Indeed, phenazone and its derivatives sold by Hoechst (now Aventis} boosted the revenues of this company; similarly Bayer profited both, aspirin and phenacetin (Brune, 1997), Bayer and other companies tried to find more drugs with an even better ellect/side-elfect profile: one was paracelamol (synthesized by Bayer). From the begin ining, it was suspected to he a metabolite af phenace lin, Bayer resorted to the advice of 3. von Mering (1893) (Fig. 2), 2 famous pharmacclogist in Stras- bourg. He got the impression that paracetamol wes not superior to phenacetin: rather, he observed that this compound retained some of the negative features of accisnilide, namely methemoglobinsemia (von Mering, 1893). This turned out later to be an error, possibly mediated by impurities, but it fostered the the first hall widespread use of phenacet ol the 20th century. ‘The situation changed in 1946 when American sei ted the phenacetin meiabolism. Lester and Greenberg at Yale together with Flinn, Brodie and Axeliod (Axelrod, L988) (Fig. 3), in New York, showed that paracetamol was the main (and active) metabolite of phenacetin, a claim originally put forward. by Treupel (1894). Brodie believed that paracetamol is the safer and more active hroughor Hinsberg and 956. tut an 192015) 955-565 Brave etal drug as compared to phenacetin, as phenacetin could lead 10 aniline, but paracetamol not (#linn and lrodie 1948). Stil, paracetamol wat bolic cascade. 1 not the end of the meta inated as glucuronide or sullate, but some paracetamol was oxidized to NAPQL (N-acetyl-p-benzoquinone imine). a highly reactive compound, which we will have to deal with later: Anyway, from the data accrued by the Axelrod group, it was plausible thar paracetamol (in the United States, avelaminephen) could be the better drug, tree of bur retaining the full analgesiciantipyretic power of phenacetin. Regretia bly, this cc yelusion was partially false These observations supported the market introduc. tion of paracelamol/acetaminophen, Interestingly. the comparatively small problem of methemoglobin pro duction was remedied, but the liver toxicity of parac elamol was overlooked until 1966 (Davidson and Eastham, 1966). The meiabolization of phenacetin to pparacctamol is slow, allowing for the detoxification of mest of the emerging paracetamol (Fig. 1) into harm: less glucuronides or sullates, Only small amounts of the toxic, reactive NAPQLare produced [rom phenace: tin (Fig. 1). The introduction of paracetamol led to a re-evaluation of the toxicity of phenacetin, This was a very timely as 2 frightening observation in Switzerland hinted at serious kidney problems from the intensive use of phenacetin 1.4 Phenacetin, the culprit of the first side-effect epidemy? Relieve of pain and fever was achieved by the use of Aspirin (introduced 1899), phenazone (1884) and acetanilide, ster substituted by phenacetin (1887). dominated the Grug market during the first 50 years of the 20th century, They were the most widely used and pre: scribed drugs in most countries (and they still are). All of them were found by serendipity, often under false assumptions. The attempts for further improvement Jed 10 the use of pharmacological analogues. Tn the United States, United Ki ind other derivates of salicylic acid were the standard analgesics. In continental Europe and parts of Asia, phenazone and derivates (aminophenazone, dipyron) were the prevailing drugs. When the major metabolite of phenacetin, paracetamol, was introduced (1953), lew consumers switched to the new compound, but the ner synthetic, antipyretic analgesics igdom and Australia, aspirin the situation changed soon Finily, the economic situation of Europe and the United States improved and the det goods with it. In Switzerland, the traditional watch {and for consumeri srune eta mand Uc reduces riser 1 Figure HU 2olinge, 5. Moes Dubach covered the an lessen, ui 1988 an wns ro: Kantonsptal Bal, CH, 1971 1912-1989 sntunt, New mechanical watches dustry gained mon were assembled under time constraint Gob work) mostly by women, ‘The job work led to headache and exhaustion after a few hours of assembling small and delicate mechani- cal_ watches (Moeschlin, 1957, 1958; D 1975, 1991). An effective method to overcome pain and fatigue was a break, calfee and a piece of bread covered with butter and punched headache pills’. Most prominent was a brand pro- duced by Hofmann-La Roche, called Saridon®. (At that time, Saridon® contained calfeine, phenacetin, propyphenazone and barbiturates (Zollinger, 1955). Acthe same time, it was customary in Switzerland that elderly women had their collee in one of the nice colfee shops. They supplemented their coffee with a couple of headache pills, which they could from the waiter. These and other habits, net only in ibach etal 0 the butter et directly Switzerland, which typically consisted of the three Ingredients: phenacetin (or phenazone, propyphenazone), aspirin and caffeine. Nobody was afraid of any harm as all the components and the mixtures of them were regarded sale, uselul and pleasant —a major error as turned out Tater. is the merit of scientists from Switzerland (Fig. 4) who observed an increase in incidence of interstitial nephritis in the population of Basel and nearby regions. They described the pathology of this ‘disease’ (better: chronic intoxication). 1 occurred olten in young (30-50 years old) women (many workers in the wateh industry) who suffered initially from mild hypertension, later from interstitial nephritis going along with pigmentation of the skin and early death from cardiac infarction and/or cardiac insulficiency. spahler and Zollinger (1953) speculated abour a the environment, including the widespread use of barbiturates and antibiotics (Spiller and Zollinger, 1953). Moreover, they realized that in most yereased the use of headache pills, dd Zollinger Hans Ulrich 130 years of aniline analgesics 4 Moeschlin Sven 1910-2005 Dubach Ulrich C. 1926-2011 ‘eases the disease started with headache, which was at first episodic and Jater elmost permanent, Today. we call it “rebound headache’ (Rapoport etal,, 1996) Many of the patients were sullering fram this type of hronie headache aud in Saridon®) regularly ‘uadistinetion to what is written in many atticles, in the original paper, Sper and Zollinger (1953) did not suspect 2 causal relationship between the analgesics and interstitial nephritis, although one= third of their (first) 44 patients had consumed analy sies regularly. I was the merit of a dlinidan, s. Moeseblin (Fig. 4}, to ping “Saridon’-abuse and kidney decay (Moeschlin, 1957, and quoted in ‘Der Spiegel’, July 1958). It was only shortly after their publication that a worldwide increase of interstitial nephritis was observed and a connection between the use of analgesics (mostly combinations) and nephropathy (analgesic nepltropa- thy, phenacetin nephritis, Saradan nephritis) was pos tulated (Nanra, 1980; Kincaid-Smith, 1981: Dubach eval, 1991) The observation from Switzerland and many other countries indicated that phenacetin was the major culprit of interstitial nephritis, hypertension and death. Consequently, beginning in the 1950s of the last century, paracctamol supplanted phenacetin. It was hoped that the abuse potential would be lower and the kidney toxicity absent, There was indeed some reduction of the ontinous interstitial neplarlis, ur the coincidence of the reduction may reflect the intensive attempts, on the one hand, to hinder consumers to abuse combination analgesics and, on the other hand, 1 get those who were habitually consuming these compounds, decondi- tioned and detoxifled (Drukker et al., 1986; Feinstein etal, 2000), It turned out again that combination analgesics taken regularly not only lost their anti- Iieadache activity but rather induced headache ested headache pills (often vwidhout much relieve. In cor jnt the relation between rather 97130 years of aniline enatgesis attacks (Rapoport etal, 1996), Tn the end, it appeared that kidney problems were solved and paracetamol could be used as a sale and active analgesic, even in pregnant women and chil- ren, Regrettably, this assumption was wrong ‘again. Paracetamol does not cause methemaglobin pro: duction (Prescott, 1980) but it turned out to be hep. lotoxic due to its very reactive intermediate (NAPQL, Fig. 1), which is produced in the liver and kidney cells via the CYP P450 metabolizing en7ymes, preferentially in the liver and kidney. Hundreds of death cases and thousands of hospital referrals result from overdosing, voluntarily and accidentally every year (Moynihan, 2002; Lee, 2004), 2. The scientific evaluation late or too late? 2.1 Kidney and liver toxi for paracetamol /: specific Following the ban of phenacetin, paracetamol/aceta~ minophen gained increasing market shares. The mar= keting claim that paracetamol does not go along wit the usual unwanted drug eliecis of the competitor aspirin was convincing. In later years, the non steroidal anti-inflammatory drugs (NSAIDs) were introduced, They cansed gastrointestinal (GI) toxicity, kidney damage, bleeds, asthma-like reactions and car- diovascular problems, including hypertension, cardiac Invfaretlons and cardiac insuificieney In Jong-term use in high anti-inflammatory doses. They looked thers fore less acceptable than paracetamol in low analgesic doses, The putative better tolerability of paracetamol lucthered sales and use. Moreover, since paracetaruol lowered fever without obvious acute side effects safety claims were widely accepted without epidemi logical proof. Negative clinical reports were rare and serious events were dedated the result of overdosing. Paracetamol was the recommended drug for pregnant women and newborn children as well as for aged and frail patients suttering, ¢.g., from osteoarthritis (OA) (Zhang et al. 2010). Dipyron, an altemative to parac- etamol, was sliown to cause ~ very rarely — agranulo: eytosis (Kaufman et al. 2006) and therefore banned in many countries. Aspirin, contrary, may cause Reye's syndrome, leading to the termination of the Worldwide use of children’s aspirin (Eder ct al., 2006). Regrettably, this ban of aspirin has started au intensive tase of paracetamol in ehildren and by that possibly a worldwide epidemy of asthma, attention-deficit/ 958 curs Pan a8 2uIs) as. mss Brune etal hyperactivity disorder (ADHD) and retarded child development (see later) 2.2 Inhibition of prostaglandin synthesis: paracetamol an NSAID? It vas only recent science that shed new light on the ‘molecular basis of effects and side elfects of paraceia mol. Several rescarch groups could disentangle the mode of action of the weak analgesic paracetamol Vane and colleagues found that paracetamal blocks prostaglandin production in brain (Flower and Vane, 1972). Using the ex vie whole blood assay (Tacconelli etal,, 2002), we and others could show that paracetamol inhibits prostaglandin produciton by cyclooxygenase (COX) not only in the central nervous system but also in the peripheral tissues of human beings (Hinz ct al., 2008). Indeed, as would be expected from a COX Inhibtioy the use of paracetamol goes along with the same prob= lems as NSAIDs. It increases blood pressure (Forman tal, 2005; Sudano et al., 2010), aggravates the risk fof GI ulcers (Rahme et al, 2008) and myocardial Infarction (Chan etal, 2006). Inhibition of blood coagulation may eccur particularly when given intra- venously at relatively high doses (for review, see Hinz and Brune, 2012), Also, paracetamol has 2 low pro- pensiiy for precipitating so-called aspirin-induced (pseudo-allergie) reactions, including asthma and shoek (Szezeklik et sl., 2003), Moreaver, the elaim that paracetamol can be used in the final stages of pregnancy because it does not inhibit prostaglandin produciion ts not justified: Recent evidence Indicates that early closure of the ductus arteriosus — a typical consequence of inhibited prostaglandin production ~ can be provoked by paracetamol as well as by ibupro Jen (Hammerman et al,, 2012). Thuis proves again that paracetamol is a COX inhibitor, which can cause clfects rypical for this type of dr adminis: tering paracetamol has become an accepted procedure for actively dosing the ductus atterivsus in pretem newborns (Hammerman et al., 2012). On the con trary, paracetamol as all COX inhibitors may conirib- tte to the (rare) event of a premature (i wfer0) closure of the ductus (Simbi et al., 2002), Moreover, as otter COX inhibitors, paracetamol can precipitate pre eclampsia (Rebordosa et al,, 2010), These are rare events. But then, paracetamol is used in low, offen inelfective doses. Prostaglandin synthesis Inhibition, however, does not explain the liver damage (including an increase of transaminases), which may occur even at permitted doses in apparently healthy consumers (Watkins et al,, 2006; Walton, 2014).K Brune etal 2.3 Liver and kidney toxicity: NAPQI Paracetamol is the most widely used analgesic world- ‘wide, but itis also the predominant reason for ALF in industrialized countries. including United States (Lee. 2004; Larson et al,, 2005), United Kingdom (Tanne, 2006) and Scandinavian connteies (Wel ct al., 2007), In the European Community, it appears to be the prevailing reason for ALF requesting liver transplan- tation (Gulmez et al., 2013). Moreover. the rate of ALT transplantation was more than twoleld higher for no-overdose paracetamol exposure than for all NSAIDs pooled. In addition, paracetamol wes invelved in 97% of cases (IT1 of 114), where transplantation vas attributed to drug use (Gulmer et al, 2013) The normal detoxification mechanisms, i.e, gli curonidation and sulfation (Fig. 1), allow for some paracetamol to be translormed into the reactive NAPOT by the CYP P450 system (particularly CYP 2E1 and 3A4; Laine et al, 2009). NAPQI reacts with sulf- hydryl groups, eg. in glutathione (GSH). GSH conju: gation is generally considered t0 be a detoxification mechanism. Adm Neacetyleysteine (NAC) restores GSH levels and protecis from liver damage, whereas GSH-depleting agents enhance par- acetamel hepatotoxicity (Mitchell et al,, 1973), In contrast, protein adduct formation with NAPAT is a mechanism of paracetamol toxicity (Follow etal, 1973), Protein adducts can be measured in the liver and in serum. NAPQI may, in e., very slim under- nourished persons or alcoholics, cause acute liver decay and death from standard doses (Moore etal. 2013), Even healthy young volunteers ingesting stan dard doses of paraceiamol show an increased release of transaminases — a sign of liver toxicity (Watkins et al., 2006). This liver toxicity was not seen with phenacetin, Indeed, several researchers wondered it by eliminating phenacetin from human use, ‘the wrong pig’ was killed Paracetamol averdase also induces renal toxicity (Curry etal, 1982; Mazer and Perrone, 2008). Metabolites from paracelamol-GS conjugates have been implicated in toxicity. Paracetay cysteine conjugates aggravated paracetamol-induced renal injury (Trumper et al, 1996; Stern et al., 2005), Recent experimental findings provide evidence lor a signalling role of NAPOI produced in murine lungs following administration of low non-toxic doses of paracetamol (Nassini et al,, 2010). NAPQI activated the transient receptor potential ankyrin-1 (TRPAL) expressed by primary sensory neurons, where it stimulated the release of the pro-inflammatory neu: ropepiides. Paracetamol induced neurogenic inflam- 130 year of aniline analgesics ‘mation in the murine lung, This effect was absent in paracetamol-treated TRPAI-deficient mice. Therelore, NAPQT may also act as a signalling molecule contrib luting to the asthma-inducing effect of paracetamol observed in. several epidemiological studies (lor review, see Henderson and Shaheen, 2013), Paracetamol is eliminated quickly [yp,:~1 h, in ehtl- dren 1 (newborn) ~ 8h]. This made it difficult to provide proof that a patient suffcring from acute liver decay has ingested paracetamol (James et al., 2006) This situation changed when several research groups could show that paracetamol is bound covalently to qsicine molecules in dillerent proteins. These acctaminophen/protein adducts remain in the body for several days. They can be detected with chromato- graphic techniques (Madsen etal... 2007). Conse- quently, the proof that paracetamol las contributed to, an acute liver damage has become easier (James et al. 2006, 2013; Heard et al,, 2011; McGill and Jaeschke, 2013). On the contrary, this technique used in healthy patients demonstrates that even at norinal duses par- acelamol adducts are produced, which, in tum, requires the production of NAPQT that obviously is not completely detoxified by conjugation with ghuia- thione, The adducts e! paracetamol allow to speculate about the contribution of paracetamol ingestion 0 otherwise unexplained liver damage. tnéced, it could bbe shown that in many unexplained intoxications in children (James et al, 2006) and adults (Heard et al 2011; Khandelwal ctal., 2011), paracetamol may have played a major role, ay even days after hospital admission protein adducts were found. Moreover, these protein/paracetamol adducts may be produced in embryos and newboms necessary enzymes are active in the fettis and paracetamol crosses the placenta barrier freely (Rollins et al., 1979; Byer et al,, 1982; Heard et al., 2011). These observa tions gain importance when scen in context with the multitude of epidernielogical and experimental results indicating long-term problems in children who were exposed {0 paracetamol (and NAPQL) during pregnancy since the 2.4 Paracetamol for pregnant women and newborn children? Paracetatnol has lost the glamour of being well toler ‘ated and innocuous to pregnant women and newbom children. Large analyses of databanks and case-control studies show that the use of paracetamol duzing preg- nancy may increase the incidence of asthma in chi dren of mothers having used paracetamol during pregnancy (for review, see McBride, 2011). Ibuprofen, furs Pan #812015) 952-965 959130 years of aniline analgesics investigated head to head with paracetamol in a pro- spective study, was not associated with an increased incidence af asthma in contrast to paracetamol (Lesko tal. 2002). In addition, investigations of the large databanks ol Scandinavian and other cou cate that paracetamol during pregnancy may (2) impair male fertility Jensen et al., 2010; Snijder etal. 2012), (2) hinder psychosocial development (Rrandlistuen etal., 2013), (G3) increase the incidence of ADHD in later life (Liew etal. 2014). ‘These observations have to be taken as signals of risk: Paracelamol use during pregnancy appears to interfere with the normal development of the newborn. They show less physical and psychological competence at the aye of 3 years as compared to siblings from the same mother who had not heen exposed to paracela- mol during this pregnancy or were exposed to ibupro- {en (Brandlistuen ct al,, 2013). Moreover, preliminary, evidence indicates that male children born from mothers taking paracetamol during pregnancy are :more prone to suller rom cryplorchidisin as compared to boys bom without having been exposed to the drug (Jensen et al., 2010) It is so surprising that different impairments of child development and health may be due to a simple dr as paracetamol, Moreover, paracetamol has been in ‘mass consumption for decades, Why were these puta- tive problems discovered only during the last years ‘with the help ol “new diagnostic tools and huge data- bases’? At a first glance, one may wonder if these problems were not ‘created’ 10 allow for ‘paracetamol bashing’. For decades, paracetamol was the only rec: ommended analgesic for use during pregnancy and during breastfeeding, Since alternatives were lacking, an increase in the incidence would have been difficult to be detected, particularly as ll ellects may happen it untreated children as well - but more seldom. Regeet ably, the helief stl provalls that paracetamol Is effee- live and safe. It gained this position without thorough toxicological investigation, lis status is based upon the general believe and not on scientific data, It is forn- hate that the Food and Drug Administration (EDA) is Irying to improve the situation by more and more wamings, reducing the recommended dosing and to eliminate analgesic combinations at least for children. It appears that only the availability of large data: banks containing information about many drugs and diseases used during and alier pregnancy, including the allspring, allow for detecting a measurable Increase of common, but not too rare problems. OL course, epidemiology has so far only produced evi- fies indi- 960 car ron 19 UDI 953-955 Brune ra dence of specific risk conditions (sec, eg, Cooper etal, 2014). The question remains as to how reliable these observations areand what the molecular mecha nism of such long-term effects could be. More prospective cohort studies (similar to that of Liew et al, 2014) should be pursued to add final proof (of the observed risks or, on the contrary. could lead to the rejection of the present claims. Even more con vineing would he prospective randomized controlled tials (RCTS). They are, however, not feasible for ethical reasons. [Further details of the mechanisms contributing to these (purported) toxic effects in preg, naney and the second generation can be found In a 1, 2014).1 recent review (Tieg 2.5 Experimental toxicology suggests that multiple toxicity is unavoidable This being said, we might resort to additional experi- mental proof of our assumptions. There are indeed experimental results that support the claim that par- acetamol duc to its metabolization and mode of action zmay be of special toxicity in pregnaney in contradic- tion to what we have told to our students and cob leagues during the last decades: ‘© Until recently, many scientists believed that the reactive metabolite NAPOT is produced by CYP p450 enzymes and active only if other metabolic pathways are overloaded oF not operating (Prescott, 1980). This is not the case, Rather, there is evidence that parac cclamol is metabolized in the tissue of the mother and Ictus to NAPOL as the fetus is endowed with the nec- essary enzymes (Schueiz et al, 1993; James et al, 2013) and paracetamol can cross the placenta freely (Byer et al,, 1982; Weigand et al., 1984) ‘+ The discovery of paracetamol protein adduets and the ability to measure their conceniration in plasma (ames et al, 2006) anil tissues (Roberts et al, 1991) allows fur the conclusion that NAPO! is not only pro- duced but also reacts whth Importani maceomalecuiles in the fetus, Depending upon the developmental phase fof the fetus or embryo, major damage may occur, leading to functional impairment during later life * On this background, one may speculate that an apairment of the development of the nervous system, 1s observed in mice exposed to paracctamol (Viber etal, 2014), may play a role in the observed retarda- tion of the development of children (Brandlistuen etal. 2013). ‘The changes in the maturation of the immunological system of the fetus may cause an increased propensity to develop asthma, as immunological stem cells may not be programmed in an adequate manner (Thiele ‘apis euresean an Fesnn Icrune eral tal, 2013) and thus contribute to the increased inci- dence of asthma alter the termination of aspirin use in children and pregnant women (Eder et al., 2006) Finally, the exposure of fertilized zebrafish eggs and larvae to NAPQI from paracetamol causes serious damage to the developing fish embryos (Weigt et al 2010). 3. Paracetamol in the future 3.1 Alternatives So far, it has always been argued that paracetamol is a weak but harmless analgesic with low costs and low risks, This characterization cannot be sustained anymore as recent epidemiological research indicates (Roberts etal, 2014, under review). Also, the claim that eliminating paracetamol from the market would deprive many patients of pain relief cannot be accepted, as less toxic (e.g., Doherty et al., 2011) COX inhibitors as diclofenac, ibuprofen, ketoprofen and naproxen — can take aver in most instances, Some physicians, however, worry that the ‘mild’ side effects ol paracetamol may be traded in for those serious ones seen with modern NSAIDs. That this is not the case in a prospective RCT in OA patients, Woherty etal, 2011). Indeed, ibuprofen was mare d Jess toxic than paracetamol alone Paracetamol fs By no means harmless. It is the only -the-counter (OTC) drug leading reliably to death at averdose (this is not the case with NSAIDs). Even the contention that it is safer as compared to its mother substance phenacetin fas never been proven and the production of small amounts of the toxic metabolite NAPOI cannot be avoided. Indeed, in retrospective, the phenacetin-comaining, analgesics may have activated two toxie mechanisms On the one hand, the kidney protection by prostagian: dins produced by the ailing kidney is inhibited by all components of these analgesic mixtures, except cat Ieine. In addition, some combinations ~ and regretia- bly, these are the ones used most widely - contain paracetamol (or formerly phettacetin), which are both iransiormed into the toxic metabolite, NAPQL In the kidney, Inadequate prostaglandin production plus NAPQT adducis in the kidney may conjointly lead 10 nephropathy causing hypertension, cardiovascular events including cardiac infarction and stroke, i symptoms that had already been observed by Spi and Zollinger (1933), Moeschlin (1957) and Dubach etal. (1975) and which ore regarded typical. for NSAIDs, In the liver, NAPQT alone may do the job. comes out clear effective 130 years of aniline analgesics 3.2 Aniline analgesics: is there a future? After using aniline derivatives in pain therapy for 130 years, we have to face the fact that hesidessmall thera peutic ellects, al these compounds are endowed with high toxicity. All attempts to reduce this toxicity by reducing the dese, substituting one aniline derivative to another or, finally, adding an antidote to the active ingredient (eg,, NAC), do not lead to a satisiyingly sale medication; Lowering the dose causes dissatisfied, patients and increases the risk of everdasing with analgesic combinations (compare recent FDA recom ‘mendations). Substituting paracetamol for phenacetin ‘has shifted the target of toxicity trom the kidney to the liver and, a8 we leart recently, to the pancreas (Cavanaugh and Nout, 2014). Adding aceryleysteine goes along with additional risks (Bateman and Dear, 2009). 3.3 Paracetamol in at-risk groups If paracetamol unavoidably carries the risk of toxicity, are there patient groups who benefit especially from paracetamol? Formerly, all experts including myself recommended paracetamol for use in specific groups within the population tat might be specifically endangered to drug-induced damage 3.3.1 Elderly It was and is stil recommended as first choice in eldeny, e.g, sulfering from OA (Zhang et al., 2010). Most patients, however, prefer a modern NSAID (Towheed et al,, 2006). Since we know that paraceia- mel as ibuprofen increases the risk of GI side elfects (Moore et al., 1999; Rampal et al., 2002), cardiovas- cular problems, pseudo-allergie attacks, etc., this post tion carmot be held anymore. This view is supported by Doherty et al. (2011) who showed in an RCT that paracetamol was more toxic and less effective than ibuprofen in OA. The recent tendency 10 use NSAIDs in low doses in elderly and children appears to be @ ‘more meaningful approach. Moreover, a recent RCT demonstrated that paracetamol does ot work in low back pain (Williams etal, 2014). These findings appear to change the view of many experts who ne Jonger suggest paracetamol as first choice in OA ar low ack pain (MecAlindon et al., 2014). 3.3.2 Childrenipregnant women Most experts recommended paracetamol during pregnaney and in children. In most countries, parac- Fe 3Pan 19.0015 953-965 961