Inflammopharmacology

, Volume 17, Issue 6, pp 275–342

Ibuprofen: pharmacology, efficacy and safety

Authors

    • Emeritus Professor of Biomedical Sciences, Biomedical Research CentreSheffield Hallam University
Review

DOI: 10.1007/s10787-009-0016-x

Cite this article as:
Rainsford, K.D. Inflammopharmacol (2009) 17: 275. doi:10.1007/s10787-009-0016-x

Abstract

Objectives

This review attempts to bring together information from a large number of recent studies on the clinical uses, safety and pharmacological properties of ibuprofen. Ibuprofen is widely used in many countries for the relief of symptoms of pain, inflammation and fever. The evidence for modes of action of ibuprofen are considered in relation to its actions in controlling inflammation, pain and fever, as well as the adverse effects of the drug.

Summary of outcomes

At low doses (800–1,200 mg day−1) which in many countries are approved for non-prescription (over-the-counter) sale ibuprofen has a good safety profile comparable with paracetamol. Its analgesic activity is linked to its anti-inflammatory effects and is related to reduction in the ex vivo production in blood of cyclo-oxygenase (COX)-1 and COX-2 derived prostanoids. Higher prescription doses (circa 1,800–2,400 mg day−1) are employed long-term for the treatment of rheumatic and other more severe musculo-skeletal conditions. Recent evidence from large-scale clinical trials with the newer coxibs, where ibuprofen was as a comparator, have confirmed earlier studies which have shown that ibuprofen has comparable therapeutic benefits with coxibs and other NSAIDs. For long-term usage (6+ months) there are greater numbers of drop-outs due to reduced effectiveness of therapy, a feature which is common with NSAIDs. Spontaneous reports of adverse events and adverse drug reactions (ADRs) in clinical trails from long-term coxib comparator studies, as well as in epidemiological studies, shows that ibuprofen has relatively low risks for gastro-intestinal (GI), hepato-renal and other, rarer, ADRs compared with other NSAIDs and coxibs. A slightly higher risk of cardiovascular (CV) events has been reported in some, but not all studies, but the risks are in general lower than with some coxibs and diclofenac. The possibility that ibuprofen may interfere with the anti-platelet effects of aspirin, though arguably of low grade or significance, has given rise to caution on its use in patients that are at risk for CV conditions that take aspirin for preventing these conditions. Paediatric use of ibuprofen is reviewed and the main results are that the drug is relatively safe and effective as a treatment of acute pain and fever. It is probably more effective than paracetamol as an antipyretic.

Conclusions

This assessment of the safety and benefits of ibuprofen can be summarized thus: (1) Ibuprofen at OTC doses has low possibilities of serious GI events, and little prospect of developing renal and associated CV events. Ibuprofen OTC does not represent a risk for developing liver injury especially the irreversible liver damage observed with paracetamol and the occasional liver reactions from aspirin. (2) The pharmacokinetic properties of ibuprofen, especially the short plasma half-life of elimination, lack of development of pathologically related metabolites (e.g. covalent modification of liver proteins by the quinine–imine metabolite of paracetamol or irreversible acetylation of biomolecules by aspirin) are support for the view that these pharmacokinetic and notably metabolic effects of ibuprofen favour its low toxic potential. (3) The multiple actions of ibuprofen in controlling inflammation combine with moderate inhibition of COX-1 and COX-2 and low residence time of the drug in the body may account for the low GI, CV and renal risks from ibuprofen, especially at OTC doses.

Keywords

IbuprofenArthritisPainAnti-inflammatoryAnalgesiaAntipyresisProstaglandinsNon-prostaglandin mechanismsAdverse reactions

Abbreviations

ADME

Absorption, distribution, metabolism and elimination

AEs

Adverse events

AUC

Area-under-the curve (plasma concentration)

CV

Cardiovascular

CoA

Coenzyme A

COX-1

Cyclo-oxygenase-1

COX-2

Cyclo-oxygenase-2

Coxibs

COX-2 selective NSAIDs (a sub-group of NSAIDs)

Cl/F

Clearance (fractional)

Cmax

Maximal plasma concentration

CNS

Central nervous system

Cp

Plasma concentration

CYP

Cytochrome P450

GI

Gastrointestinal

Kel

Elimination rate

NOS

Nitric oxide synthase (iNOS, inducible eNOS, endothelial, nNOS, neuronal)

NSAIDs

Non-steroidal anti-inflammatory drugs

NS-NSAIDs

Non-selective NSAIDs, i.e. those with approximately equivalent COX-1 and COX-2 inhibitory effects

OR

Odds ratios

OTC

Over-the-counter, non-prescription

PMNs

Polymorphonuclear leucocyte

Rac-

Racemic

tNSAIDs

traditional NSAIDs, e.g. aspirin, ibuprofen, naproxen

t1/2

Half-time of elimination

Tmax

Time of maximal concentration

UDP

Uridine diphosphate

VD

Volume of distribution

Introduction

Ibuprofen is one of the most widely used analgesic–antipyretic–anti-inflammatory drugs today. It probably ranks after aspirin and paracetamol in non-prescription over-the-counter (OTC) use for the relief of symptoms of acute pain, inflammation and fever, although the patterns of use of these analgesics vary considerably from country to country. Of these three analgesics, OTC ibuprofen is probably the least toxic, being rarely associated with deaths from accidental or deliberate ingestion or with serious adverse reactions. Indeed, it has been described as “the mildest NSAID with the fewest side effects which has been in clinical use for a long time” (General Practice Notebook, http://www.gpnotebook.co.uk, accessed 12/11/07).

Ibuprofen was initially introduced in the UK in 1969 and afterwards during the 1970s worldwide as a prescription-only medication, where it was recommended to be prescribed at up to 2,400 mg day−1 (or higher dose in the USA) for the treatment of musculo-skeletal pain and inflammation as well as other painful conditions (Rainsford 1999a). In the 1970s it was often prescribed either as a first line NSAID or in place of aspirin, indomethacin or phenylbutazone for treatment of arthritic conditions where it had a reputation for good efficacy and lower gastrointestinal adverse effects. Initially the drug was used in low doses ranging from 400 to 1,200 mg day−1 and with latitude and experience by physicians’ cautious dose-escalation proceeded to the current recommended dosage of 2,400 mg day−1. The emphasis on cautious use of ibuprofen was one of the hallmarks of its early success and the increasing confidence that it was safe (Rainsford 1999a).

Over the years there have been many challenges to ibuprofen, some from concerns about safety including the occurrence of some very rare but serious adverse reactions [e.g. Stevens-Johnson and Lyell’s (toxic epidermal necrolysis) syndromes, renal or hepatic failure, necrotising fasciitis] as well as some that are more common to the class of NSAIDs (Khan and Styrt 1997; Zerr et al. 1999; Rainsford 2003). Most recent of these have been cardiovascular (CV) conditions that were highlighted by the occurrence of myocardial infarction and cardio-renal symptoms in patients receiving the newer class of NSAIDs, the coxibs [rofecoxib, valdecoxib and to some extent celecoxib (Ostor and Hazleman 2005; Psaty and Furberg 2005; Rainsford 2005a; Topol 2004, 2005)]. This had the effect of regulatory agencies worldwide examining the potential of all NSAIDs, to cause CV and cardio-renal symptoms, an aspect that is still of concern for some of the coxibs and some other NSAIDs.

There have also been many challenges from newer NSAIDs, particularly the wave of some 20–30 new NSAIDs introduced in the period of 1970s–1980s and the much publicised introduction of the selective cyclooxygenase-2 (COX-2) inhibitors (coxibs) that appeared in 1999 following the discovery of COX-2 as the main prostaglandin synthesizing enzyme expressed in inflammation and pain pathways (Rainsford 2007). Surprisingly, over half of the NSAIDs introduced to the clinic since the 1970s have been withdrawn mostly due to unacceptable and unpredictable toxicities. In a sense, ibuprofen has survived these challenges both from the point of view of competition from the newer drugs and the inevitable negative impact of the failures of other drugs and associated safety issues (e.g. the CV risks raised by the coxibs). Mostly, these issues have concerned prescription-only NSAIDs although those sold OTC like ibuprofen may also have been affected by these issues.

Another major competitor has been paracetamol (acetaminophen) especially in the OTC field (Peterson 2005) but also, as discussed later, in the therapy of osteoarthritis. In non-prescription OTC, paediatric use both ibuprofen and paracetamol are equally effective in controlling fever but there are recent data to suggest that combination of these two drugs may be particularly useful in severe febrile or painful conditions (Hay et al. 2006, 2009). Arguably, ibuprofen and paracetamol have differing modes of action (see later sections) and so it is possible that they may have additive or even synergistic effects.

Claims by those advocating paracetamol is that its use is associated with lower gastrointestinal (GI) and renal adverse reactions than observed with ibuprofen. For OTC use these differences are minimal or nonexistent (Rainsford et al. 1997; Bjarnason 2007; Moore 2007). At higher doses used in arthritis therapy the consensus is that the differences in GI adverse reactions are marginal, while renal adverse reactions may be more prevalent in patients taking ibuprofen. A major issue with paracetamol is hepatotoxicity especially when taken in the range of 3–4 g daily long-term and with alcohol. The situation about consumption of alcohol with paracetamol and the use of this drug in patients with alcoholic liver disease or signs of alcohol abuse is quite serious. These aspects have been considered extensively elsewhere (Novak and Lewis 2003; Larson et al. 2005; Myers et al. 2007; Suzuki et al. 2009).

Naproxen and ketoprofen have been competitors with ibuprofen, in those countries (e.g. USA, UK and Australia) where they are marketed for OTC use. Both these drugs are more potent as anti-inflammatory agents and prostaglandin inhibitors than ibuprofen and are associated with higher risk of upper GI adverse reactions at prescription doses (Hersh et al. 2000a, 2007; Milsom et al. 2002). Naproxen tends to be used as a second-line drug for treatment of primary dysmenorrhoea where aspirin, ibuprofen and paracetamol are found less effective. Ketoprofen is favoured by some for more severe joint pain in arthritic disease.

Overall, ibuprofen has withstood competition and challenges over the 4 decades since its introduction as a prescription drug and over 2 decades since it was introduced for OTC sale.

Oral pharmacokinetics

Ibuprofen (R/S) in the form that is present in most OTC ibuprofen-containing preparations as well as the generic preparations sold by prescription (for daily doses up to 2,400 mg) and non-prescription or OTC sale (for daily doses up to 1,200 mg) exists as a diastereoisomeric mixture. This comprises half as the S(+) enantiomer which is pharmacologically active as a prostaglandin (PG) synthesis inhibitor and the other half mass as R(−) ibuprofen which is less active as a PG synthesis inhibitor but which may have some pharmacological properties relevant to the anti-inflammatory actions of ibuprofen (Brocks and Jamali 1999; Rainsford 1999b, 2003; Graham and Williams 2004). About 40–60% of the R(−)-form of ibuprofen is metabolically converted to the S(+) form (Fig. 1) (Rudy et al. 1991; Brocks and Jamali 1999) in the intestinal tract and liver after oral absorption (Jeffrey et al. 1991; Jamali et al. 1992).

The first step involves the activation of R(−)-ibuprofen with ATP Mg to form the AMP-derivative which is then esterified with coenzyme A by the action of acyl-CoA synthetase. The R-ibuprofen-CoA undergoes epimerization via the actions of epimerase to form S-ibuprofen-CoA which is then hydrolysed by a hydrolase to form S-ibuprofen (Brocks and Jamali 1999) (Fig. 1).
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig1_HTML.gif
Fig. 1

Conversion of R(−)-ibuprofen to its S(+) enantiomer via catalytic activity of fatty acyl coenzyme thioesterase. From Graham and Williams (2004) reproduced with permission of the publishers, Taylor & Francis, London

Pharmacokinetics in adults

Ibuprofen is rapidly absorbed from the upper GI tract with peak values of the R(−) and S(+) drug in the plasma or serum at approximately 1–2 h (Brocks and Jamali 1999; Graham and Williams, 2004; Table 1) with some variations according to pharmaceutical formulation (Jamali et al. 1988; Aiba et al. 1999; Halsas et al. 1999; Higton 1999; Scott et al. 1999; Trocóniz et al. 2000; Ding et al. 2007). The plasma S/R ratio can vary according to the time-release characteristics of the formulation with higher ratios being obtained with sustained-release (SR) compared with immediate-release (IR) formulations, respectively (Ding et al. 2007). A liquid or liquigel formulation of ibuprofen has proven to be popular for rapid analgesia which is related to fast absorption (Laska et al. 1986; Seymour et al. 1991; Hersh et al. 2000b).
Table 1

Pharmacokinetic properties of oral ibuprofen in adults

Dose of Ibuprofen

R(−)-ibuprofen

S(+)-ibuprofen

Reference

Cmax (μg mL−1)

Tmax (h)

t1/2 (h)

AUC (μg h mL−1)

CI/F (L h)

VD (L)

Cmax (μg mL−1)

Tmax (h)

t1/2 (h)

AUC (μg h mL−1)

CI/F (L h)

VD (L)

400 mg

15.4

1.0

1.6

44.5

  

16.2

1.0

2.4

67.4

  

Hynninen et al. (2006), Cheng (1994)

400 mg

  

1.74

 

3.52

7.8

23.5

3.0

1.7

100.8

3.4

4.6

Ding et al. (2007)

600 mg (IR)

20.8

2.96

1.2

65.9

5.0

6.45

23.9

 

2.4

90.8

  

Jamali et al. (1988)

600 mg III

20.3

 

2.0

56.9

  

26.7

 

2.2

100.2

   

600 mg IV

17.3

 

2.2

51.5

  

15.6

2.0

2.6

63.9

  

Tornio et al. (2007)

400 mg

16.4

1.5

2.9

50.1

  

19.0

1.64

2.18

3.3

75.0

[286.8]

  

Suri et al. (1997)

Chen and Chen (1995)

400 mg

17.8

1.59

1.33

52.2

4.0

7.46

26.0

1.3

2.0

93

1.02

 

Cox et al. (1991)

800 mg (Boots)

  

3.1

23.9

2.6

11.0

16.8

1.9

2.2

32.4

  

Geisslinger et al. (1989)

800 mg (Motrin®) (Arthritic pts.)

24

1.1

1.8

67

1.52

 

16.4

1.8

2.3

91.7

  

Geisslinger et al. (1990)

600 mg

14.8

1.7

1.8

57.2

5.37

17.8

20.3

2.4

2.1

86.2

13.3

 

Geisslinger et al. (1993)

600 mg

14.3

1.5

2.3

54.3

5.66

18.7

11.3

1.13

1.9

43.7

  

Evans et al. (1990)

600 mg (arthritic pts.)

 

2.3

2.0

67.6

14.0

 

16.9

1.87

3.4

75.4

  

Evans et al. (1990)

200 mg

10

1.1

1.9

23.9

4.19

11.2

31.6

1.89

2.0

141

  

Evans et al. (1990)

400 mg

14.1

1.4

3.2

41.8

4.78

22.3

47.7

2.19

2.0

216

  

Evans et al. (1990)

800 mg

24.4

1.6

2.1

73.6

5.44

16.7

6.1

1.0

 

19.5

  

Jamali and Kunz-Dober (1999)

1,200 mg

29.5

1.6

4.2

86.1

6.97

42.6

14.5

1.0

 

45.5

   

200 mg (pre-surgery)

6.1

1.0

 

17.7

  

8.7

1.1

 

31.6

   

600 mg

14.8

1.0

 

35.3

  

16.9

1.63

2.65

70.4

   

Mean (SD)

13.1

2.02

2.15

78.2

   

200 mg

8.1

1.1

 

20.8

  

28.8

1.6

2.4

117

   

400 mg

15.4

1.3

2.7

47.2

4.1

12.5

       

600 mg

17.1

1.9

1.7

55.5

4.0

14.3

       

800 mg

24.2

1.4

2.3

70.0

3.2

13.9

       

Table 1 summarizes the pharmacokinetic properties of R/S-ibuprofen, the data for which has been derived from various studies in volunteers and patients with arthritic conditions.

As can be seen from the data in Table 1 the mean (±SD) values for many of these parameters in adults show remarkable consistency from the different studies and indicate that ibuprofen has, in general, predictable and reliable kinetic properties. Furthermore, there are dose-related plasma concentration, Cp, and to some extent AUC values but the kinetic constants reflected by t1/2 (or the inverse, Kel) suggest that there is little variation with dosage. There is also little variation of these kinetic parameters with repeated dosage.

The pathways of oxidative metabolism of ibuprofen are shown in Fig. 2. These principally involve cytochrome P450 2C9 (CYP-2C9), CYP-2C8 and 2C19 participating in the oxidation of the alkyl side chain to hydroxyl and carboxyl derivatives.
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig2_HTML.gif
Fig. 2

Oxidative metabolism of ibuprofen. (From Graham and Williams, 2004; reproduced with permission of the publishers, Taylor & Francis, London)

Phase 1 Metabolism of R(−) and S(+) ibuprofen involves hydroxylation of the isobutyl chains to 2 or 3-hydroxy-derivatives and subsequent oxidation to 3-carboxy-ibuprofen and p-carboxy-2-propionate; these oxidative reactions being catalyzed by cytochromes P450 2C9 (CYP-2C9) and CYP-2C8 (Fig. 2, Graham and Williams 2004). There appears to be differential involvement of these cytochrome isoforms on the metabolism of the enantiomers with CYP-2C9 favouring formation of S(+)-2 and S(+)-2 hydroxy-ibuprofen and CYP-2C8 favouring R(−)-2-hydroxyibuprofen formation (Hamman et al. 1997). Inhibition of CYP-2C8 by administration of gemfibrozil to humans increases the plasma concentrations of R(−)-ibuprofen by about one third along with prolonging the elimination half lives of R(−) and S(+) by 54 and 34%, respectively, and increase of AUC values by about 20% (Tornio et al. 2007). This suggests that CYP-2C8 plays a major role in oxidative metabolism of the ibuprofen enantiomers. Differences in genotypes of CYP-2C9 have been associated with marked variations in drugs that are metabolized by this isoform (Kirchheiner and Brockmöller 2005). Thus, compared with CYP-2C9*1/*1 individuals with either *1/*2, *2/*2 *1/*3 or *3/*3 variants have 88, 78, 72 or 55% reduction in the clearance of the drug, respectively (Kirchheiner and Brockmöller 2005). In Spanish populations the occurrence of varying allelic frequencies in CYP-2C*8 and the CYP2C*3 allele have been shown to result in decreased metabolism of ibuprofen leading to increased AUC0→∞ and reduced clearance (López-Rodriguez et al. 2008). For other NSAIDs (e.g. celecoxib, diclofenac) there is either increased or decreased clearance in individuals with these isoforms. So there is marked variation in the pharmacokinetics of ibuprofen and other NSAIDs according to the CYP-2C9 and CYP2C8 status.

Differential metabolism of S(+) and R(−)-ibuprofen occurs by the CYP-2C9 and CYP-2C8 with these being referred to as S(+) ibuprofen and R(−)-ibuprofen hydroxylase activities, respectively (Kirchheiner et al. 2002). The allelic frequencies of these CYP isoenzymes the 3 ascribed to CYP-2C9 comprise the wild type CYP-2C9*1 which is characterized by an arginine at codon 359 on the gene. In the variant CYP-2C9*2 this arginine is replace by cysteine, and in the variant CYP-2C9*3 the isoleucine-359 is replaced by leucine. In vitro studies and human PK studies have shown that CYP-2C9*2 has only slightly less activity than that of the wild type CYP2C9*1 whereas that of CYP-2C9*3 is 10–30% less so (Kirchheiner et al. 2002). In comparisons of the pharmacokinetics of the S(+) enantiomer the rates of clearance were found to parallel the enzymic activity with subjects having the CYP-2C9*1/*2 and *3/*3 variants having 27 and 53% less clearance than those with the wild type *1/*1 genotype (Kirchheiner et al. 2002).

Other studies have examined the role of CYP-2C9 and CYP-2C19 polymorphisms for associations with drug-induced idiosyncratic reactions (Pachkoria et al. 2007). While arguably liver reactions from NSAIDs may be associated with abnormalities of phase 1 and phase 2 metabolism, the studies by Pachkoria et al. (2007) have failed to establish if polymorphisms of CYP-2C9 or CYP-2C19 are associated with liver disease.

There do not appear to be any differences in R(−)/S(+) pharmacokinetics with sex in adults (Knights et al. 1995; Walker and Carmody 1998) with the exception of the volume of distribution VD/F being about twice that in adult females compared with males (Walker and Carmody 1998). As noted later there are age differences in plasma levels and kinetics of ibuprofen; elderly subjects have slightly prolonged values of elimination half life for the total drug concentrations (Albert and Gernaat 1984) or the AUC for S(+) ibuprofen (Chen and Chen 1995).

Compromised liver metabolism in patients with moderate to severe cirrhosis leads to prolongation of the t1/2 to 3.1 h and 3.4 h for R(−) and S(+) ibuprofen with evidence of reduced metabolic inversion of the R(−) to S(+) enantiomer (Li et al. 1993). Alcoholic liver disease also prolongs the Tmax for the total Cp as well as the t1/2 (Albert and Gernaat 1984). Elevation of the AUC for S(+) ibuprofen and higher S(+)/R(−) ratios have been observed in patients with a variety of chronic inflammatory diseases who have impaired renal function (Chen and Chen 1995).

Surgical removal of wisdom teeth has been found to cause substantial reduction in the serum concentrations of both R(−) and S(+)-ibuprofen (by 2.6 and 3.5-fold, respectively) with prolonged Tmax from approximately 1 h to 4–6 h depending on the dose (Jamali and Kunz-Dober 1999). It was suggested that a number of stress-related factors influencing GI functions could contribute to reduced gastric absorption of ibuprofen (Jamali and Kunz-Dober 1999). These observations are of particular therapeutic importance for they argue in favour of higher doses of the drug for pre-emptive or post-operative surgery.

Phase II metabolism involves formation of phenolic and acyl glucuronides (Rudy et al. 1991; Kepp et al. 1997; Brocks and Jamali 1999; Graham and Hicks 2004) and a minor route of conjugation with taurine which is stereospecific to the S(+) enantiomer because of formation from the thioester CoA which participates in the R(−) to S(+) conversion (Shirley et al. 1994). Biliary excretion in humans of unchanged drug and active phase II metabolites accounts for about 1% of the drug, which compares with 50% of the urinary excretion (Schneider et al. 1990). The 15 known UDP-gluronyltransferases that catalyze formation of glucuronides in human liver have been shown to be controlled by 5 UGTIA and 5 UGT2B genes and the development of these proceeds from birth to 6 months of age (Strassburg et al. 2002). Hepatic glucuronidation of ibuprofen has been found to be 24-fold lower in children aged 13–24 months than in adults (Strassburg et al. 2002). This low level of detoxification of ibuprofen is clearly an important consideration for safe dosage of the drug to infants.

Stereospecific disposition of ibuprofen enantiomers occurs into the synovial fluids of arthritic patients, many of whom have synovitis or inflammation of their knees. There is appreciable accumulation of R/S-ibuprofen in synovial fluids with broad peaks occurring over a period of 2–6 h which follows the peak plasma or serum concentrations (Glass and Swannell 1978; Mäkelä et al. 1981; Albert and Gernaat 1984; Gallo et al. 1986). The ratios of total ibuprofen concentrations in the synovial fluid to those in plasma is about 1.24 at 7 h following single dose of 600 mg of the drug and 0.52–1.46 at 3–12 h after 3 daily doses of ibuprofen 1.8 g day−1 (Gallo et al. 1986). The mean free total ibuprofen in synovial fluid ranges from 1.81 to 2.91% compared with that in plasma which is 1.54–2.53%. Thus, there is appreciable total and free R/S-ibuprofen that accumulates in synovial fluids of arthritic patients and clearly this will have therapeutic significance in relation to the local anti-inflammatory and analgesic effects of the drug in pain control. In studies of the disposition of the individual enantiomers it has been found that the concentrations of the S(+) isomer as well as values of AUC S(+) always exceed those of the R(−) enantiomer (Day et al. 1988; Cox et al. 1991; Geisslinger et al. 1993; Seideman et al. 1994) with similar selective accumulation being shown in experimentally induced skin suction blisters (Seideman et al. 1994). The patterns of synovial fluid accumulation of the enantiomers follows that of the peak plasma levels with broad peaks of R(−) and S(+) ibuprofen at about 2–4 h and extending to about 12–15 h (Seideman et al. 1994) thus showing persistence of the enantiomers in synovial fluids well past those of the peak plasma concentrations of these enantiomers.

As with many other NSAIDs inter-subject variability is a common feature of the pharmacokinetics of ibuprofen (Wagener and Vögtle-Junkert 1996) which does not relate to variation in the rates of inversion of R(−) to S(+) ibuprofen (Geisslinger et al. 1993). The kinetics of the enantiomers appears similar after multiple compared with single dosage of the racemate (Cox et al. 1991). Synovial fluid concentrations of the enantiomers vary less than those in plasma and plasma values after about 5.5 h; it being inferred from pharmacokinetic analysis that the accumulation is primarily of protein-bound drug (Day et al. 1988).

Non-bound (to albumin) NSAID concentrations (i.e. free) are generally considered to be those which are pharmacologically relevant to the actions of these drugs as well as being of relevance to the untoward effects of drug–drug interactions where toxic effects of NSAIDs or other drugs relate to displacement of one or other following binding to albumin or other plasma proteins. As with many NSAIDs most of which bind to plasma proteins to around 99%, ibuprofen also is strongly bound to albumin (Brocks and Jamali 1999; Graham and Hicks 2004).

The binding of ibuprofen in plasma compared with synovial fluids depends on the concentrations of albumin in these compartments (Wanwimolruk et al. 1983). There are differences in the affinity of S(+) compared with R(−) ibuprofen for binding on human serum albumin; there being two binding sites for both ibuprofen enantiomers of varying affinity (Hage et al. 1995; Itoh et al. 1997) with some evidence of allosteric cooperative binding at high concentrations of S(+) ibuprofen (Hage et al. 1995). There may be greater binding of S(+) ibuprofen to the site II (diazepam) binding site of albumin (Cheruvallath et al. 1997).

Ibuprofen accumulates in the cerebrospinal fluid (CSF) of patients (undergoing lumbar puncture for treatment of nerve root compression) with the AUC’s of the R and S enantiomers in the CFS being 0.9 and 1.5% those in plasma, respectively (Bannwarth et al. 1995). The estimated t1/2’s were 3.9 and 7.9 h for the R(−) and S(+) enantiomers (compared with 1.7 and 2.5 h in plasma). The pathological condition being treated in these patients involves neuro-inflammatory reactions with associated drug accumulation in the central nervous system, so these observations may have relevance to central analgesic actions of ibuprofen.

Two other routes of metabolism of ibuprofen which though they do not appear to have clear pharmacological significance have been considered to be of theoretical significance. Thus, the stereoselective uptake of R(−) ibuprofen forming mixed hybrid triglycerides as a result of thioester-CoA formation and subsequent fatty acid metabolism (Williams et al. 1986) has been considered important though wanting for evidence for its significance either toxicologically or pharmacologically.

The acylation of proteins to form adducts from reaction of acyl glucuronides of ibuprofen (Castillo et al. 1995; Vandenhoeven et al. 2006) may be of relevance for adverse reactions in the liver analogous to those occurring with some other NSAIDs though evidence for this is lacking.

In summary, the pharmacokinetic properties of ibuprofen (Brocks and Jamali 1999; Graham and Williams 2004) can be summarized thus:
  1. (1)

    There are relatively fast rates of absorption of the drug with subsequent “first pass” liver phase 1 and phase 2 metabolism to well-characterized (a) phenolic and carboxylic acid derivatives via CYP-2C8, CYP-2C9 and possibly CYP-2C19 activities, and (b) conjugates with glucuronic acid and taurine (a minor metabolite).

     
  2. (2)

    The biodisposition of ibuprofen reflects high plasma protein binding and low volume of distribution but with the capacity to be accumulated in appreciable quantities in inflamed compartments where there is need for anti-inflammatory/analgesic activity (synovial fluids, CSF).

     
  3. (3)

    Ibuprofen has a relatively short plasma elimination half-life and although prolonged in liver and renal diseases this is not so appreciable as to be a factor accounting for higher frequency of adverse events in patients with these conditions compared with those with relatively normal hepato-renal functions. Indeed, the short plasma half-life (t1/2) has been suggested as a factor accounting for relatively low incidence of serious GI events compared with traditional NSAIDs (bleeding, peptic ulcers) (Henry et al. 1996, 1998).

     
  4. (4)

    Ibuprofen exhibits approximately linear kinetics to within 1,200 mg dosage or near compliance with expected kinetics. Thus, when taken up to 1,200 mg repeated doses of ibuprofen the elimination is not saturated.

     
  5. (5)

    Chronic disease states (arthritis) have relatively little impact on the overall kinetics of ibuprofen. However, acute surgical pain reduces the plasma concentrations of R(−) and S(+)-ibuprofen which may arise from the stressful conditions of the surgery. This has been suggested as evidence to necessitate considering dosage adjustment in the therapy of acute surgical pain on the basis of allowance for increasing dosage to meet adequate pain control.

     
  6. (6)

    The t1/2, AUC, VD and clearance kinetics of conventional ibuprofen tablets suggest that the usual dosage regime of either 400 mg t.i.d. for OTC use or 400–800 mg t.i.d. or q.i.d. as appropriate for prescription use to 2,400 mg daily. Extended-release formulations that have been developed could enable twice daily dosage to limits of 1,200 mg day−1 OTC or 2,400 mg day−1 prescription requirements.

     

Pharmacokinetics in children

Before considering some of the factors affecting the PK of ibuprofen in young adults and children it is useful to consider some of the differences in drug metabolism and disposition in these groups. Physiological development in neonates, infants and children has considerable impact on the absorption, distribution, metabolism and elimination (ADME) of drugs (Kearns and Reed 1989). Among the notable differences in drug disposition between infants and children compared with neonates and young adults are the decline in total body water (TBW), increase in intracellular water (ICW) and reduced extracellular water (ECW) that occurs post-natally up to about 1–2 years with subsequent variable increases in TBW and ICW up to about 25 years where upon those in males are greater than in females (Kearns and Reed 1989). Likewise, fat content increases post-natally up to about 1 year, plateaus to mid-teens then declining until mid 20s where upon differences occur between the sexes with females having greater body fat than males (Kearns and Reed 1989). These changes in body composition along with plasma protein concentrations will have marked effects on the volume of distribution of drugs. Protein binding of drugs is influenced by a variety of factors and the total concentration of plasma proteins is decreased in the neonate and infant compared with that in children (Kearns and Reed 1989).

Age-related changes in hepatic biotransformation are evident from the neonatal period where they are not developed and approximate adult values by about 6 months (Kearns and Reed 1989). Thus, Phase I hepatic transformation progressively increases to near adult values by 6 months of age (Kearns and Reed 1989; Kearns 1993). Glucuronidation increases from ±10 days to 2 months while conjugation with amino acids increases from ±10 days to 3 months (Kearns and Reed 1989; Strassburg et al. 2002). These physiological factors have considerable significance in the age-related pharmacokinetics of paracetamol and NSAIDs in children (Walson and Mortensen 1989; Jacqz-Aigrain and Anderson 2006). Changes (usually as increased values) of t1/2, VD, and C of these drugs occur progressively from the neonatal period to 1–3 years, whereupon they often (but not always) assume adult values. Of particular relevance is the possibility that the febrile state which is often treated with analgesics/NSAIDs can influence drug metabolism and biodisposition as a consequence of release of cytokines [e.g. interleukin-1 (IL-1), tumour necrosis factor-α (TNF-α)] during infections which are pyrogenic and alter drug metabolism.

The pharmacokinetics and what is known of the pharmacodynamic properties of ibuprofen in <12 year children can be considered to be similar to that of young-middle aged adults in whom most investigations have been performed. A possible caveat to this application of adult pharmacokinetics to that in children/infants (<12 year) might be related to differences in growth rates thus affecting body mass and gender affecting hormonal regulation of drug metabolizing enzymes.

With the exception of the conversion of R(−)-ibuprofen to its S(+) enantiomer the pharmacokinetic parameters of ibuprofen in children are comparable with those in adults (Table 1 c.f. Table 2). It appears that the rate of conversion of R(−) ibuprofen is lower in children than it adults (Table 2). In a study in 11 infants (6–18 months) the plasma levels of the S(+) enantiomer of ibuprofen were lower than in adults while the values for t1/2 for R(−) and S(+) ibuprofen are within the range of those expected in older children or adults (Kauffman and Nelson 1992; see also Jacqz-Aigrain and Anderson 2006). It is suggested that the relatively low levels of S(+) ibuprofen argue for a higher dosage of ibuprofen in infants.
Table 2

Pharmacokinetics of Ibuprofen in children

Oral absorption

t1/2: 0.3–0.9 h

Tmax: 1–2 h

10 mg kg−1 → Cmax: 44 mg L−1

Protein binding

99%

Active isomer

S(+)

Plasma concentration

S(+) children < adults

Metabolism

CYP450 2C9 and 2C8

t1/2

0.9–2.3 h

From Autret-Leca (2003)

Other appreciable changes in paediatric populations have been observed in young children aged less than 5 years where the clearance (CL/F) and volume of distribution (VD/F) may be less than that in adults or older children and the plasma half-life of elimination (t1/2) prolonged to about twice that in adults or older children (Jacqz-Aigrain and Anderson 2006) (Table 3).

Given these provisos it should be possible to describe the pharmacokinetic (PK) and pharmacodynamic (PD) properties of ibuprofen in <12 year olds as in general being related to the information that is available in adults. A few PK studies have been performed in children that overlap the 12–18 year age group suffice to conclude that, in general, the PK properties are similar to those in adults. Less is known about PD properties in the young (12–18 year) except that dose-related pain relief is similar in young adults to that in younger children and also in older adults. This suggests that the PD properties of ibuprofen shown in studies in adult populations, and supported by in vitro as well as animal model investigations are likely to be similar in all age groups.

Some pharmacokinetic parameters for ibuprofen in children of various ages are shown in Table 3. In essence, the major overall features that vary in children compared with young to mid-aged adults are (1) the greater proportion of the S(+)-enantiomer compared with the R(−)-form in plasma, and (2) the greater variation in t1/2 (Autret-Leca 2003; Jacqz-Aigrain and Anderson 2006). The VD for S(+) ibuprofen in children is greater than in adults (Kelley et al. 1992) and may reflect a higher unbound fraction in plasma compared with that of R(−) ibuprofen (Brocks and Jamali 1999).
Table 3

Pharmacokinetic parameter estimates for NSAIDs in paediatric patients modified from Jacqz-Aigrain and Anderson (2006)

Age

Formulation

CL/F (ml h−1 kg−1)

V/F (L kg−1)

t1/2 (h)

Ibuprofen

 22–31 weeksa

iv

2.06 (0.33)

0.062 (0.004)

30.5

 28.6 (1.9) weeksa

iv

9.49 (6.82)

0.357 (0.121)

43.1 (26.1)

 0.5–1.5 years

Suspension

110 (40)

0.20 (0.09)

1.6 (0.4)

 11 months–11 years

Suspension

57.6

0.164

1.97

 3 months–12 years

Suspension

80 (10) SE

0.16 (0.02) SE

1.44 (0.15)

 3 months–12 years

Suspension

110 (10) SE

0.22 (0.02) SE

1.37 (0.09)

 5.2 (1.7) years

Suspension

140 (32)

0.27 (0.11)

1.4 (0.5)

 5.2 (2.5) years

Tablet

114 (26)

0.26 (0.1)

1.6 (0.4)

 4–16 yearsb

Suspension/granules

71 (CV 24%)

(4.05 L h−1 × 70 kg)−1

Vc 0.06, Vp 0.1

(CV 65%)

Variability presented as SD in parentheses, range (x–y) or SE

CL/F, apparent drug plasma clearance; iv, intravenous; t1/2, elimination half-life; V/F, apparent volume of distribution; Vss volume of distribution at steady state; Vc, initial volume of distribution; Vp, apparent volume of distribution of peripheral compartment

aAge is gestation age (GA, weeks)

bData reported using allometric model. Estimate presented for a 30 kg individual

Inspection of Table 3 shows that the t1/2 and VD of ibuprofen in patients receiving i.v. drugs is about 25-fold higher than from orally-administered ibuprofen yet there is the same order of elimination and distribution of oral ibuprofen from an early age of about 0.5 year; thereafter the t1/2 and VD are within the range of that in adults. The rates of clearance are, however, greater in young children up to about 5 years and decline in higher age groups and are appreciably lower in i.v. administered infants (Table 3). Ibuprofen has lower glomerular filtration in premature infants and this may be a factor accounting for higher t1/2 and VD in this group compared with that in adults.

Rectal pharmacokinetics

Fundamental physiological and pharmaceutical considerations in drug administration by the rectal route

Drugs in suppositories administered by the rectal route are placed in intimate contact with the rectal mucosa which is pH 7.2–7.4 and has a lipoidal barrier (Florence and Attwood 1998). Suppositories are in contact with the mucous membrane of the rectal ampulla which comprises a layer of epithelial cells without villi (Florence and Attwood 1998). The main blood supply to the rectum is in the superior rectal or haemorrhoidal artery while drug absorption takes place through the venous network of the submucous plexus which then becomes the inferior, middle with superior rectal veins. The latter two veins connect to the portal veins and thus transport drugs direct to the liver. The inferior veins enter the inferior vena cava and thus by-pass the liver. The proportion of drug that is absorbed by these two venous routes depends on the extent to which the suppository migrates in its original or molten form up the intestinal tract. Thus, this can be variable and so drugs administered rectally may not bypass the liver (Florence and Attwood 1998).

The factors influencing rectal absorption of drugs include (a) the melting point and liquefaction properties of the suppository, and (b) solubility properties of the drug that initially influence contact of the drug with the mucosa. Aqueous solubility and pKa of the drug influence absorption from “fat” based or liposoluble drugs. Viscosity of the base and excipients or dispersants added to disperse the fat can influence absorption. The rate-limiting step in drug absorption for suppositories made from a fatty base is the partitioning of the dissolved drug from the molten base, not the solubilization of drug in body fluids (Florence and Attwood 1998).

NSAIDs and paracetamol vary considerably in their rates of absorption when administered rectally (Van Hoogdalem et al. 1991). The formulations of these drugs clearly are a major factor in influencing their absorption. For example addition of increasing amounts of lecithin can delay the rectal absorption of diclofenac (van Hoogdalem et al. 1991). The physico-chemical properties of NSAIDs can influence their absorption. Thus, ketoprofen is well-absorbed with bioavailability approaching that of the orally administered drug (van Hoogdalem et al. 1991).

Basic studies with rectal ibuprofen

Amongst the early studies were those performed by Eller et al. (1989) in which eight healthy volunteers received solution or suspension formulations of ibuprofen for administration rectally in a randomized fashion. The bioavailability for these forms was compared with that of the orally administered drug. In essence, the results showed that both rectal formulations showed similar extent of bioavailability being about 60% of the oral formulation; the Cmax values being 62–67% and the Tmax was longer than with the oral formulation.

In their studies, Eller et al. (1989) compared the bioavailability of rectally or orally administered sodium or aluminium salts of ibuprofen as aqueous buffered solutions (pH 7.8) or suspensions (pH 5.2) in eight normal healthy, non-obese, male subjects using a Latin square design. The rectal solutions were administered as retention enemas following prior treatment the night before with PhosphoSoda® then 1 h before a Fleet enema. The orally administered preparations were taken following an overnight fast. Both the rectally administered preparations had significantly less bioavailable as shown by the AUC values (Table 4) and were relatively high as shown by the Cmax values compared with the respective oral solutions/suspensions. However, as expected, the Tmax values were longer for the rectally administered preparation than those taken orally (Table 4). The t1/2 were almost identical for the oral and rectal solutions and about 1/3 lower with the oral suspension compared with the former or the rectally administered suspension.
Table 4

Bioavailability of rectal compared with oral solutions/suspensions of ibuprofen in eight normal, non-obese male human volunteers

Parameter

Treatment A

Oral solution

Treatment B

Oral suspension

Treatment C

Rectal solution

Treatment D

Rectal suspension

Peak concentration (μg mL−1)

80.7 (6)

28.7 (28)

50.3 (36)

19.2 (63)

AUC (μg mL−1 h)

 0–12

197.8 (12)

164.2 (21)

172.5 (36)

97.6 (73)

 0–∞

200.3 (12)

179.1 (30)

175.5 (36)

102.9 (74)

Peak time (h)

0.33 (30)

2.12 (28)

1.14 (36)

2.44 (45)

Mean residence time (h)

2.60 (12)

5.99 (23)

3.19 (6)

4.49 (24)

Terminal elimination rate constant (h−1)

0.351 (11)

0.211 (19)

0.344 (6)

0.367 (22)

From Eller et al. (1989)

The rectal solution had greater bioavailability than the suspension and achieved higher serum Cmax values than the suspension (Table 4). Likewise, the mean residence time (MRT) was shorter for the rectal solution than the suspension. These basic results showed that the sodium solution was the preferred salt to be used in any fundamental considerations of suppository formulations.

Glowka (2000) studied the enantiomeric pharmacokinetics in rabbits of suppositories of ibuprofen acid and the lysine salt prepared in the lipophilic base, Witepsol H-15 and showed that there was no presystemic inversion of R(−) to the S(+) enantiomers. The S:R ratios only increased after about 1.5 h following administration of both formulations and were greater with the lysine salt. The values of AUC were greater after administering ibuprofen acid suppositories compared with the lysine salt even though the latter were more rapidly absorbed.

Studies with suppository formulations

Kyllönen et al. (2005) investigated the R(−) and S(+) pharmacokinetics of Burana® (Orion Pharma, Espoo, Finland) suppositories of ibuprofen in 9 full-term infants aged 1–7weeks, 8 infants aged 8–25weeks, 7 infants aged 26–52weeks and 7 adults aged 20–40 years following administration of approximately 19–20 mg kg−1 ibuprofen suppository after induction of anaesthesia for minor general or orthopaedic surgery in infants or lumbar disc surgery in adults.

The results summarized in Table 5 show that in all age groups ibuprofen was rapidly absorbed from the suppository formulation with the Tmax in infants for R/S ibuprofen being 1.6–3.3 h and the t1/2 for absorption (the so-called “physiological” half-life) being 1.9–2.9 h. In four of the youngest group of infants (1–7weeks, Group 1) the Tmax was similar to that in those where the suppository was retained even though the Cmax values were about 40% less than in the non-retained suppository group. The only differences in Tmax for R/S ibuprofen were observed in the adults (group 4) where this was 3.3 h and so was greater than in all the other groups (infants) which ranged from 1.6 to 1.9 h.
Table 5

Pharmacokinetic variables of the R(−), S(+) and R/S ibuprofen in infants, children and adults that received suppositories of 20 mg kg−1 racemic ibuprofen perioperatively

 

(S)-(+)-ibuprofen

(R)-(−)-ibuprofen

(R, S)-(±)-ibuprofen

AUC ratio

Group 1 (n = 5) suppository retained

 Cmax (mg L−1)

29.3 ± 16.2

23.8 ± 9.4

49.2 ± 20.7

 

 Tmax (h)

2.2 ± 1.0

1.8 ± 1.3

1.9 ± 1.2

 

 Chronological t1/2 (h)

2.9 ± 1.8

3.2 ± 2.7

4.6 ± 5.1

 

 Physiological t1/2 (h)

5.8 ± 3.5

6.6 ± 5.4

8.9 ± 10.1

 

 AUC (mg L−1 h)

159 ± 81*

112 ± 54

299 ± 69*

1.7 ± 1.1

Group 1 (n = 4) suppository expelled

 Cmax (mg L−1)

12.4 ± 6.4

13.4 ± 8.1

25.7 ± 14.2

 

 Cmax (h)

1.9 ± 0.9

1.9 ± 0.9

1.9 ± 0.9

 

 Chronological t1/2 (h)

3.8 ± 2.9

3.1 ± 2.4

2.9 ± 2.1

 

 Physiological t1/2 (h)

7.8 ± 5.8

6.3 ± 5.1

6.0 ± 4.4

 

 AUC (mg L−1 h)

66 ± 40

54 ± 48

108 ± 83

1.6 ± 1.4

Group 2 (n = 8)

 Cmax (mg L−1)

38.5 ± 20.7

40.0 ± 21.8

75.6 ± 44.6

 

 Tmax (h)

1.6 ± 0.7

1.4 ± 0.8

1.6 ± 0.7

 

 Chronological t1/2 (h)

1.7 ± 0.4

2.2 ± 0.7

1.9 ± 0.5

 

 Physiological t1/2 (h)

3.1 ± 0.9

3.9 ± 1.4

3.4 ± 1.0

 

 AUC (mg L−1 h)

131 ± 79

124 ± 67

248 ± 153

1.1 ± 0.2

Group 3 (n = 7)

 Cmax (mg L−1)

42.7 ± 16.0

49.7 ± 23.3

87.9 ± 36.6

 

 Tmax (h)

1.7 ± 0.3

1.6 ± 0.7

1.6 ± 0.3

 

 Chronological t1/2 (h)

2.8 ± 1.3

1.8 ± 0.4

2.1 ± 0.7

 

 Physiological t1/2 (h)

4.6 ± 2.3

2.9 ± 0.7

3.6 ± 1.3

 

 AUC (mg L−1 h)

180 ± 98

167 ± 56

339 ± 136

1.1 ± 0.4

Group 4 (n = 7)

 Cmax (mg L−1)

30.1 ± 12.5

30.1 ± 9.9

63.8 ± 20.4

 

 Tmax (h)

3.5 ± 0.8

2.9 ± 1.0

3.3 ± 0.8

 

 Chronological t1/2 (h)

2.1 ± 0.3

2.5 ± 0.7

2.2 ± 0.4

 

 Physiological t1/2 (h)

2.1 ± 0.3

2.5 ± 0.6

2.2 ± 0.4

 

 AUC (mg L−1 h)

160 ± 65

177 ± 59

334 ± 123

0.9 ± 0.1

From Kyllönen et al. (2005), reproduced with permission

Values are mean ± SD. Only those patients in group 1 in whom the suppository was retained were included in the comparisons between the groups 1 and 4

AUC ratio is the ratio of (S)-(+)-ibuprofen AUC to that of (R)-(−)-ibuprofen

* Significantly (P < 0.05) different from the corresponding value in group 1 where the suppository was expelled

 Significantly (P < 0.05) different from the corresponding value in group 4

Surprisingly, the AUC values for the R/S, R(−) and S(+) isomers were similar in all the groups except, as expected in the youngest infant group who had expelled suppositories.

The values of Tmax for R(−) ibuprofen ranged from 1.4 to 2.9 h, with the higher value (2.9 h) in adults in contrast to the range of values in infants (1.4–1.8 h) and in which there was no significant difference between the infant groups but there was between the two older infant groups and the adult group. Only in adults was the Tmax greater for the S(+) isomer (3.5 h) than for the R(−) enantiomers (1.6–2.2 h). Since the R(−)/S(+) ratios of AUC values were greater in the youngest infant group (1.7 retained and 1.6 expelled suppository groups) compared with those in all the other groups (0.9–1.1) this suggests that there is a greater rate of conversion of R(−) to S(+) ibuprofen from suppositories in adults and which is similar to that observed following oral administration of the drug. The t1/2 of plasma elimination (so-called chronological half-life) of both the racemic ibuprofen as well as the R(−) and S(+) enantiomers was greater in the youngest of the infant groups compared with those in others and adults indicating slower rates of elimination in young infants in whom drug metabolising enzymes are not fully developed.

These results show that rectal administration of ibuprofen is an easy and effective way of achieving therapeutic plasma concentrations especially in the peri-operative and post-operative periods. Except for delayed absorption of ibuprofen in adults (which may have arisen because of the stress of the more extensive disc hernia surgery) (e.g. as in dental surgery, see Jamali and Kunz-Dober 1999) and higher physiological and chronological half-life in infants aged 1–7weeks there are no major differences in pharmacokinetics between infants and adults.

Aside from these investigations there do not appear to have been any other published studies on the pharmacokinetics of rectal ibuprofen in children or infants.

Pharmacokinetic aspects of importance in the actions and safety of ibuprofen

During the discovery of ibuprofen pharmacokinetic issues were of particular importance especially in the pre-clinical evaluation of the toxicity of the drug (Rainsford 1999a). The Boots Company had already experienced problems with ibufenac in causing hepatic reactions in patients with rheumatoid arthritis during early stage clinical trials. Furthermore, there was a major objective in the pre-clinical programme to discover a drug which was safer to the gastrointestinal tract than was evident with aspirin and some other NSAIDs at the time. Thus, to obviate the possibility that ibuprofen might cause hepatic reactions as previously seen with ibufenac evidence was obtained from radiobelled-drug studies to show that there was a lower rate of accumulation in the liver with ibuprofen than with ibufenac (Rainsford 1999b).

In considering the pharmacokinetic profile of ibuprofen (Tables 1 c.f. 6) there are several important features that should be noted;
Table 6

Summary of pharmacokinetic aspects of ibuprofen in relation to safety in adults

Short t1/2

 Healthy normal subjects

  R = 1.6–4.2 h

  S = 1.9–3.4 h

 OA patients (Dose dependent)

  R = 1.7–2.9 h

  S = 2.0–3.0 h

Renal clearance R/S

 Variably affected by arthritic state but to minor extent

 Increased by 50% at >70 years but age does not otherwise affect PK

 With insufficiency increase in AUC (S+) with age and hypertension

 Variability with pain response—increased renal clearance

Liver disease

 t1/2 increased; AUC (S+) increased; glucuronides decreased

Possible small effects of gender

  1. 1.

    The drug has a relatively short plasma elimination half-life (t1/2) a feature which has been identified in comparative studies of gastrointestinal gastro-ulcerogenicity (Henry et al. 1996) which is probably a key safety feature. The plasma half-life of the drug averages between 2 and 3 h with there being some inter subject and intra subject variability but not such that this vastly influences half-life values (Brocks and Jamali 1999; Graham and Williams 2004). Differences have also been observed in the bioconversion of the inactive (R−) an enantiomer to (S+) enantiomers and in their clearance under conditions of acute surgical pain.

     
  2. 2.

    There is no evidence of accumulation in elderly or retention in specific body compartments. There is no evidence of formation of bio-reactive metabolites sufficient to cause covalent modification of liver or other proteins that might contribute to toxicity of the kind seen in the case of paracetamol-induced irreversible hepatic injury (Graham and Hicks 2004). Glucuronide conjugates of ibuprofen represent the major metabolites of the drug and it has been speculated that these conjugates might lead to formation of adducts such as seen with other phenyl propionic and benzoic acid NSAIDs (Castillo et al. 1995). Whether these conjugates contribute to covalent modification of proteins that leads to toxicity is not known. There does not appear to be any indication of any appreciable accumulation of ibuprofen in the liver and other organs such as might result from such covalent modification. It is very likely that there is a considerable degree of spontaneous hydrolysis both of the glucuronides and the ibuprofenyl-derivatives of proteins.

     
  3. 3.

    There is little variation in the main PK properties of ibuprofen in different arthritic states.

     
  4. 4.

    Mild/moderate renal impairment does not appear to cause any elongation of the plasma elimination half-life and there is little evidence of alterations in plasma pharmacokinetics in patients with mild hepatic disease. Clearly, patients with considerable renal impairment or liver malfunctions should not be taking ibuprofen as there would be an expected increase in risk of systemic accumulation although this risk is probably of a low order in comparison with the short plasma half-life of the drug.

     

Paediatric pharmacokinetics in the febrile state

Among the most frequent indications for use of ibuprofen in children is for the treatment of fever. Since febrile conditions lead to elevation of febrile-inducing pro-inflammatory cytokines (especially IL-1β, TNFα, IL-6) and these can lead to alterations in the activities of drug-metabolizing enzymes it is important to understand if the pharmacokinetics of antipyretic agents is altered in febrile conditions in children. Earlier reviews (e.g. Walson and Mortensen 1989) emphasized the lack of PK data in children; a situation that has been addressed more extensively in recent years, although there are still some gaps in knowledge of PKs of antipyretic/analgesic drugs especially in infants.

Nahata et al. (1991) studied the PK of ibuprofen in 17 patients (aged 3–10 year) with fever from streptococcal pharyngitis and otitis media who received 5 and 10 mg kg−1 liquid formulation of the drug (mean ± SD ages for this group being 6.7 ± 2.5 and 6.2 ± 2.1 year.) The peak (mean ± SD) serum concentrations of the total of the racemate [i.e. R(−) and S(+)-ibuprofen] in these two groups were 28.4 ± 7.5 and 43.6 ± 18.6 μg mL−1 which were evident at 1.1 and 1.2 h, with the 5 and 10 mg kg−1 doses, respectively. The t1/2s were 1.6 h in both groups and the rates of oral clearance 1.2 ± 0.4 and 1.4 ± 0.5 mL min−1 kg−1 with the two dose levels, respectively, showing that the serum PK are unaffected by dose. An earlier study by Walson et al. (1989) using liquid ibuprofen in febrile children showed that the values of the racemic Cmax were slightly lower at the 5 mg kg−1 dose than observed in the study by Nahata et al. (1991) but were within the same range.

In a later study the same group performed a randomized, double-blind, parallel-group placebo-controlled study in 56 infants and children (0.5–12 year) who were primarily investigated for antipyretic effects (Nahata et al. 1992). They were given 5 and 10 mg kg−1 ibuprofen suspension or placebo in separate groups but blood samples for PK assay of the plasma concentrations of the racemate in only 17 patients, who received the drug alone. The mean maximal Cmax of the racemate were 28.4 and 43.6 μg mL−1 at 1.0 and 1.5 h for the 5 and 10 mg kg−1 dosage groups, respectively. These plasma values (Nahata et al. 1992) correspond closely with those in serum which were obtained in the earlier study (Nahata et al. 1991) showing consistency both in plasma c.f. serum and between the studies.

Another study in febrile patients was performed by Kauffman and Nelson (1992) in 49 infants and children aged 3 month to 10.4 year, the primary purpose being to investigate the relationship between plasma concentration of the racemic form of the drug and antipyretic effects. Fever was diagnosed as arising from a variety of conditions including pneumonia, otitis media, upper respiratory tract infection, tonsillo-pharyngitis and various other conditions. The dose of ibuprofen was 8 mg kg−1 which is between that of 5 and 10 mg kg−1 used in the earlier studies. Further discussion about the therapeutic effects of this and other studies reported in this section will be considered in the next section. However, it was found that there was a delay in peak concentration of ibuprofen and maximal decrease in temperature, highlighting that the therapeutic benefit follows the peak or optimal plasma concentrations of the drug.

The kinetic analysis showed that there was no affect of age on the pharmacokinetic properties of the drug in a final group of 38 patients. The results showed that oral ibuprofen suspension was rapidly absorbed with a Cmax of 35.8 ± 16.7 μg mL−1 (mean ± SD) at 0.7 ± 0.5 h (mean ± SD). The absorption was faster than that in earlier studies and similarly the half-life of absorption was fast (t1/2abs 0.3 ± 0.3 h). The plasma elimination t1/2 was 1.6 ± 0.7 (mean ± SD) h which was within the range observed in other studies and in adults.

Brown et al. (1992) investigated the disposition of 5 or 10 mg kg−1 ibuprofen and 12.5 mg kg−1 paracetamol in 153 febrile children. The Cmax occurred about 2½ h earlier than the maximal antipyresis with both drugs thus being in agreement with the study of Kauffman and Nelson (1992). The plasma AUC0–00 was lower for the high dose of ibuprofen than the lower, an observation which is at variance with that obtained in other studies.

In a investigation of the pharmacokinetics of the R(−) and S(+) enantiomers of ibuprofen in febrile children, Kelley et al. (1992) employed a randomized, open-label parallel study design in which 39 patients (11 month–11.5 year) received 6 mg kg−1 ibuprofen suspension or 5–10 mg kg−1 paracetamol. For some unexplained reason only values of Cmax being 33.5 ± 14.7 (mean ± SD) μg mL−1 and Tmax being 60 ± 19.7 min were recorded but not the values for the individual enantiomers.

Pharmacokinetics during analgesia

The disposition of the enantiomers was studied in 11 infants (6–18 months), who were anaesthetized for minor genitor-urinary surgery and given 7.6 ± 0.3 mg kg−1 ibuprofen oral suspension post-operatively (Re et al. 1994). The values of racemic, S(+) and R(−) were 24.4 ± 6.6, 9.7 ± 2.9 and 11.8 ± 4.4 μg mL−1 at Tmax approximately 2–4 h, respectively. It was apparent from these studies that the peak plasma concentrations were much longer than those observed in the previous studies in febrile infants and children suggesting that either the surgical-anaesthetic procedure delayed GI absorption of the drug or the age of the infants influenced the PK of ibuprofen. The lower S/R ratio obtained is in contrast to that of other investigators in infants where this was higher.

There are three paediatric groups where ibuprofen has been investigated for therapeutic benefits in relation to pharmacokinetic properties. These are for relief of pain and joint symptoms in juvenile idiopathic (chronic) or juvenile rheumatoid arthritis (JIA, JRA, respectively), the i.v. administration in patent ductus arteriosus (PDA) and in high doses in cystic fibrosis (CF). While both these treatments may be considered outside the norm none-the-less they are potentially important uses of the drug in therapy. Moreover, these studies provide important therapeutic data on the pharmacokinetic properties of ibuprofen in extremes of dosage and administration which, with safety data, are important for giving outside values for indications for the drug.

Pharmacokinetics in patients with juvenile rheumatoid (or idiopathic) arthritis

Limited studies have been undertaken to investigate the pharmacokinetics of oral ibuprofen in patients with juvenile rheumatoid arthritis (JRA; juvenile idiopathic (or chronic) arthritis, JIA). The production of pro-inflammatory cytokines and other inflammatory reactions would be expected to have profound consequences for drug metabolism and biodisposition in these patients.

From a pharmacokinetic viewpoint there are indications that alterations in serum/plasma albumin concentrations, especially if subnormal in severe disease may affect the percentage free concentrations of NSAIDs in the circulation, which might have some toxicological or therapeutic consequences (Skeith and Jamali 1991; Furst 1992; Litalien and Jacqz-Aigrain 2001).

Before the introduction of ibuprofen, aspirin was the most widely and successfully used drug for treating pain and joint symptoms in juvenile arthritis (Ansell 1973, 1983).

Adverse effects are, however, more frequent with aspirin as well as some of the other NSAIDs (indomethacin, meclofenamic acid, naproxen; Skeith and Jamali 1991) whereas these appear fewer with ibuprofen (Ansell 1983; Furst 1992).

The dose of ibuprofen employed in juvenile arthritis (30–40 mg kg−1 day−1) is much higher than generally employed in infants and children for the treatment of fever and painful conditions (5–10 mg kg−1 day−1) and is more in line with that employed in cystic fibrosis. Reference to the extensively studied PK properties of ibuprofen in CF may, therefore, be instructive and complement those in juvenile arthritis.

Among the earlier published reports of the PK of ibuprofen in juvenile arthritis were two studies by Mäkelä et al. (1979, 1981). These authors determined the concentrations of racemic drug in serum and synovial fluids in 17 patients with juvenile arthritis (aged 1.5–15 years) who received approximately 40 mg kg−1 day−1 ibuprofen. A striking feature of this study was that although there was a high degree of variability in the serum and synovial fluid concentrations that the ratio of ibuprofen in the synovial fluids was relatively high compared with that in the serum (Fig. 3). The absorption of oral ibuprofen is rapid and comparable to that in adults (Mäkelä et al. 1979, 1981). In 33 patients (1.5–15 years) that received approximately 40 mg kg−1 day−1 t.i.d., peak serum concentrations Cmax were 31 μg mL−1 at 1.0–2.0 h. The t1/2 was 2.3 h which is comparable with that in adults.
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig3_HTML.gif
Fig. 3

Reproduced with permission of the publishers of Clinical Pharmacokinetics of oral ibuprofen in juvenile rheumatoid arthritis. Redrawn from Skeith and Jamali (1991)

The relationship of serum/plasma concentrations to therapeutic efficacy appears to be best understood by comparing the concentrations of the pharmacologically active S(+) enantiomers (Skeith and Jamali 1991).

Cystic fibrosis

Though not a recognised indication for the drug it has been explored for therapy of cystic fibrosis and these studies are useful for consideration of safety and efficacy of ibuprofen. In particular, data on the PK of ibuprofen in cystic fibrosis (CF) are quite extensive and are instructive as they give information on drug disposition at high dosages, in a condition where there is considerable pulmonary (often with accompanying Pseudomonas or other bacterial infections) and gastrointestinal inflammation as well as systemic manifestations of disease some of the respiratory inflammatory symptoms of which are markedly ameliorated by ibuprofen. Thus, there is considerable interest in these data from safety aspects.

Konstan et al. (1995) have pioneered use of ibuprofen for treating CF. In a randomized, double-blind, placebo-controlled, dose-escalating study in 19 children (6–12 years) in Ohio (USA) Konstan et al. (1991) compared the plasma PK of racemic (R/S) ibuprofen based on its enantiomers (assayed by HPLC) following administration of 300 mg o.d. of the drug for the first month, 400 mg o.d. in the second month and 600 mg o.d. in the third month. The dose of ibuprofen was increased if the peak plasma level was ≤50  μg mL−1 and if placebo then this was also increased. The PK of ibuprofen was also investigated in 13 children who received 13.4 ± 4.1 mg kg−1 (mean ± SD) compared with that in 4 normal children who received similar doses of the drug.

In the former investigation 11 patients in the ibuprofen and 5 in the placebo group who completed the study underwent 24 PK studies. In the dose escalation study values of Cmax were 38, 29 and 65 μg mL−1 for the 3 dosages 300, 400 and 600 mg day−1, respectively. The Tmax values were 68, 128, 109 min indicating that at the highest dose there was some limitation to gastric absorption. Indeed, there are indications of absorption effects and a wide scattering of Cmax data in relation to dose (mg kg−1) of ibuprofen suggesting that some of the GI effects of the disease (hyper mucus secretion) may influence absorption of the drug. Compared with PK in normal adults or those with arthritic diseases (Tables 4, 5) the values of Cmax and Tmax are higher by a twofold factor or greater. The values of AUC (5.8, 6.3 and 10.8 mg min−1 mL−1) for the three doses also appear higher than in adults with the rates of clearance (1.8, 2.1, 1.9 ml min kg−1) being relatively low. The t1/2 was approximately 68, 128 and 109 hr for each dosage level reflecting extension of residence time of the drug in the body. Thus, these investigations show that there are marked differences in the PK of ibuprofen in CF patients compared with young or mid-aged adults.

In the detailed second part of this investigation the plasma concentrations and AUC values in 13 CF patients (6.1 ± 1.7; mean ± SD mg min−1 mL−1) were significantly lower than in controls (11.3 ± 3.4, mean ± SD, mg min−1 mL−1) with reduction in clearance by about 1/3 and an approximate equivalent increase in VD. The possible reasons for these substantial alterations in PK include decreased bioavailability (possibly from GI absorption), increased metabolic clearance and increased unbound fraction in plasma leading to increased clearance through the kidneys (Brocks and Jamali 1999).

In another study in North Carolina, USA in 38 children of both sexes age ranges 2–13 years with CF the enantiomer PKs were investigated in a single-dose, open-label investigation following 20 mg kg−1 racemic ibuprofen (Dong et al. 2000). The S:R enantiomeric ratio of the plasma AUC was 2.09:1 and the free and conjugated ibuprofen in urine was 13.9:1 (S:R) which indicated there were no differences in these parameters compared with those in normal children. While there were no differences observed in other PK parameters there was an inverse relationship between the CI/F for R(−) ibuprofen with age in CF patients. There was no significant difference in PK parameters with sex or formulations (suspensions, tablets) of ibuprofen.

The dose of ibuprofen employed by Doyle et al. (2002) was 20 mg kg−1 and was greater than that in the second PK study by Konstan et al. (1991) (13.4 mg kg−1 in CF and 13.9 in controls) so the differences in PKs between these studies might be explained, in part, by differences in dosages even though the actual values for the R(−) and S(+) enantiomers were not clear from the study by Konstan and co-workers.

Arranz et al. (2003) investigated the population PK of serum ibuprofen in 59 CF patients (2–18 year) in order to identify the factors accounting for inter individual variability. The PK analysis revealed that the inter-individual variability was such that the absorption constant (Ka) could not be estimated accurately. Dose-dependent kinetics was demonstrated affecting clearance and VD. As would be expected, the fasting status and formulation (acid or lysine salt) of the affected the bioavailability and clearance of ibuprofen while slower absorption of the free acid was evident compared with that of the lysine salt of ibuprofen. It should be noted that these authors did not assay for the enantiomers of the drug and so the ratio of S:R could be another factor accounting for population variability (Re et al. 1994).

Pharmacodynamic activities

Ibuprofen has multiple actions on different inflammatory pathways and cellular systems involved in acute and chronic inflammation (Rainsford, 1999b, 2003). A summary of the principle inflammatory reactions that are affected by ibuprofen is shown in Table 7. This summary is essentially based on various experimental observations. In some cases there are paradoxical actions (e.g. on the synthesis of nitric oxide (NO) which are system-dependent and do not always show uniform effects. What is presented here is a consensus view of the modes of action of ibuprofen.
Table 7

Summary of modes of action of ibuprofen and its enantiomers

Effect

Reference

Inhibition of prostaglandin synthesis

 Rac–Ibu: is an equi-potent (unselective) inhibitor of COX-1 and COX-2

Mitchell et al. (1994), Warner et al. (1999), Kawai et al. (1998), Kargman et al. (1996), Boneberg et al. (1996)

  S(+) more potent inhibitor than R(−) on COX-1 and COX-2; no time-dependent but reversible inhibition

Boneberg et al. (1996), Gierse et al. (1999)

  R(−) competes with S(+) at active site on COX-1

Rainsford (1999b, 2003)

  S(+) more potent inhibitor of COX-1 than COX-2

Boneberg et al. (1996)

  Rac-Ibu more potent inhibitor of peroxidase activity of COX-2, than COX-1

Gierse et al. (1999)

  Thoester of R(−)-Ibu and S(+) competitive inhibitors of COX-1 and COX-2

Neupert et al. (1997)

Inhibition of leukotriene production

 Ca++-ionophore-stimulated LTB4 and 5-lipoxygenase by PMN’s reduced equally by Rac−, S(+) and R(−) Ibuprofen

Villanueva et al. (1993)

Inhibition of leucocyte functions

 Reduced migration of PMN’s, expression of ICAM-1, E-Selectin and vascular adhesion molecules

Hofbauer et al. (2000), Menzel et al. (1999), Zhang et al. (2006)

 Reduced oxyradical production by fMLP stimulated PMN’s and xanthine oxidase

Chan et al. (2008), Perez et al. (2007)

 Reduced release of β-glucuronidase from fMLP-stimulated PMNs

Villanueva et al. (1993)

Nitric oxide production and actions

 Reduced exhaled NO and urinary nitrite/nitrate in human volunteers and NO in lungs of rats after endotoxin

Vandivier et al. (1999), Akbulut et al. (2005)

 Transient increase, followed by sustained decrease of NO by PMNs after oral S(+) ibuprofen 400 mg

Vural et al. (2002)

 Decreased NOS protein & mRNA in LPS and γIFN-stimulated macrophages or glial cells, but increased iNOS and NO in endothelial cells

Menzel & Kolarz (1997), Stratman et al. (1997), Aeberhard et al. (1995), Miyamoto et al. (2007)

 Decreased NO-mediated vasodilation

Hardy et al. (1998)

 Decreased NO, peroxynitrite and oxyradical production by human PMNs

Costa et al. (2006)

 Increased iNOS in stomach after ibu-arginate

Tegeder et al. (2001)

Inhibition of production of transcription factors, MAPkinase, nuclear

receptors, heat shock proteins

 Decreased NFκB, IκB, IKK, Erk 1/2, pGORSK, NAG-1, HSF-1, Hsp70, PPARα, PPARσ

Lagunas et al. (2004), Gomez et al. (2005)

 Inhibition of IL-1β-induced cAMP levels

Vij et al. (2008)

Inhibition of cytokine production

 Reduced production of IL-1β and IL-6 in microdialysates in forearm of skin after adjacent sunburn-inflammation by 800 mg but not 400 mg Ibuprofen in volunteers, coincident with pain reduction

Angst et al. (2008)

 Increased production of c-reactive protein and pro-Inflammatory cytokines in blood of ultramarathon runners following racing

Nieman et al. (2006)

 S(+) and R(−) ibuprofen equally active as inhibitors of monocyte IL-1 & TNFα

Hofbauer et al. (2000)

Inhibition of antibody (IgG, IgM) synthesis

 Reduced in human peripheral monocytes

Bancos et al. (2009)

Apoptosis

 

 Reduction of apoptosis and TNFα induced by Fas signalling in vivo

Cazanave et al. (2008)

 Reduces NO donor effects on cell viability & apoptosis

Asanuma et al. (2001)

Anandamide production

 Inhibits adadamide hydrolase in nervous system causing increased endogenous cannabinoid (anandamide)

Fowler et al. (2009)

The principle pharmacodynamic (PD) actions of ibuprofen, like that of other NSAIDs, that are involved in control of acute pain, fever and acute inflammatory reactions are the inhibition of COX-1 and COX-2 derived pro-inflammatory prostanoids (mainly PGE2) (Rainsford 1999b; Burian and Geisslinger 2005; Hinz and Brune 2006). Pain relief is attributed to peripheral (anti-inflammatory) and central nervous system (CNS) effects of S(+) ibuprofen on the inducible COX-2 and iNOS/cNOS present in inflamed or inflammatory cells of the peripherally affected regions as well as in the dorsal horn and higher spino-thalamic tracts mediating pain transmission. There may be some contribution of inhibition of COX-1 in the CNS to the analgesic actions of ibuprofen (Martínez et al. 2002). Recently, much interest has been shown in the possibility that ibuprofen may enhance the synthesis of endogenous cannabinoids and so contribute to the central analgesia via cannabinoid receptor activation (Fowler et al. 2009) as well as acting on NMDA receptors (Björkman et al. 1996). There is possibly a contribution of R(−) ibuprofen to the effects on both the leucocyte activation neural activity and spinal transmission influencing the effects of ibuprofen in inflammatory pain. In analgesic studies the pharmacokinetics of ibuprofen has been found to relate to the cortical evoked potentials and subjective pain response following tooth pulp stimulation in human volunteers (Suri et al. 1997). These observations indicate that the plasma profile of the active S(+) enantiomer which is the potent PG synthesis inhibitor parallels the analgesic response and the ex vivo inhibition of PG synthesis in the blood of volunteers (Suri et al. 1997).

Antipyretic effects of ibuprofen are principally due to the effects of S(+) ibuprofen on the synthesis of PGE2, the main signalling mediator of pyresis synthesized in the hypothalamic—preoptic region at the base of the brain; pyrogens such as bacterial lipopolysaccharides and the cytokines and cell breakdown (lytic) products produced from activated leucocytes being the source of pyrogens that activate receptors in the hypothalamic-preoptic area causing production of COX-2 derived PGE2 (Rainsford 1999b). Anti-inflammatory effects of both R(−) and S(+) ibuprofen on activated leucocytes at inflamed sites could reduce the release of pyrogenic molecules and so contribute to the anti-pyretic effects of the drug.

Ibuprofen is frequently used in control of menstrual pain in young adult women (Milsom et al. 2002; Grimes et al. 2006). The major therapeutic effects of the drug in this condition are attributed to the inhibition of PGF and PGE2 that are responsible for the uterine smooth muscle spasm and local inflammatory pain in this condition (Dawood and Khan-Dawood 2007).

Ibuprofen is also widely used in dental pain, including that from impacted third molars and in their surgical removal; these being painful conditions frequently experienced by young adults (Dionne 1998; Dionne and Cooper 1999; Beaver 2003; Derry et al. 2009). Extensive clinical investigations have shown the effectiveness of OTC dosages (circa 800–1,200 mg) of ibuprofen in pain relief in the surgical removal of third molars that is superior to placebo and in most studies that of paracetamol where it is equal to or more effective than this frequently used analgesic (Dionne and Cooper 1999; Beaver 2003; Derry et al. 2009). Recently, ibuprofen has been used as a reference drug for comparing effects of coxibs in dental pain where it has been found to be about equally effective in pain relief (Morse et al. 2006; Bannwarth and Bérenbaum 2007; Cheung et al. 2007; Geusens and Lems 2008). The analgesic effects of ibuprofen in dental surgery have been attributed to (a) inhibition of the local (dental) production of PGE2 by S(+) ibuprofen as well as the anti-oedemic activity of the parent drug (Dionne and McCullagh 1998; Dionne and Cooper 1999), and (b) increase in plasma β-endorphins (Troullos et al. 1997); paradoxically, however, results from the same group have shown the opposite effects with S(+)-ibuprofen which produced rapid analgesia (Dionne and McCullagh 1998). Studies in rats suggest that ibuprofen (as well as rofecoxib) may reduce the expression of COX-2 in dental pulp (Holt et al. 2005).

The effectiveness of ibuprofen in headache and migraine has been demonstrated in a number of studies in children and young adults (Stewart and Lipton 1998; Suthisisang et al. 2007; Silver et al. 2008). Less is known about the mechanisms of action in these conditions but it is probably due in part to S(+) ibuprofen affecting platelet activation and thromboxane A2 production and local vascular effects in the affected regions of brain vessels. It is significant that ibuprofen can penetrate into the CNS so that this may contribute to the central analgesic effects including that in headache and migraine as a consequence of local accumulation of the drug (Bannwarth et al. 1995; McCrory and Fitzgerald 2004; Kokki et al. 2007). Several experimental studies suggest that S(+) ibuprofen administered intrathecally (i.a.) into the CNS has direct analgesic effects which are greater than the R(−) form (Malmberg and Yaksh 1992; Björkman et al. 1996).

Ibuprofen at OTC doses is extensively used in sports and other acute minor injuries where it has proven effectiveness over placebo and equivalent or superior effects to some other NSAIDs and non-narcotic analgesics. In this and other acute pain conditions the OTC dosages (~1,200 mg day−1) suffice to relieve pain although occasionally higher anti-arthritic doses (up to 2,400 mg day−1) may be employed. Most pain studies show placebo effects amounting to approximately 30% reduction in VAS scores or other indices of pain measurement and the effects of ibuprofen are at least double this or greater and more prolonged (>6 h) than placebo (Kean et al. 1999; Rainsford 1999a, b; Moore 2007).

Modes of action in relation to pharmacokinetics

What has emerged since ibuprofen was initially discovered is that (a) the drug undergoes metabolism of inactive R(−)-ibuprofen enantiomer (present as half the mass of ingested drug) to the prostaglandin synthesis inhibitory or active S(+) isomer, and (b) these isomers each have pharmacological effects of importance to the therapeutic actions of the drug, thus highlighting the interrelationships between its metabolism and pharmacodynamic effects (Fig. 4) (modified from Rainsford 1999b using information in Table 7).
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig4_HTML.gif
Fig. 4

Principal metabolic routes of R(−) and S(+)-ibuprofen and actions of metabolites. Modified from Rainsford (1999b) with permission of the publishers, Taylor & Francis and updated using information from Table 7 and references therein

Like many conventional (or traditional) tNSAIDs, ibuprofen inhibits both the constitutive cyclo-oxygenase-1 (COX-1) which is responsible for production of prostanoids (PGs and thromboxane A2, TxA2) that control a range of physiological or “housekeeping” functions (vascular, blood flow, gastric and renal functions), and the inducible COX-2 whose synthesis is increased leading to amplified production of PGE2 in inflammation and pain (Cryer and Feldman 1998; Rainsford 1999b, 2004a, b; Warner et al. 1999; Vane and Botting 2001). COX-1 inhibition has been considered a factor underlying the possibility of NSAIDs to cause some adverse effects (GI ulcers, bleeding, renal abnormalities) although there are other biochemical and cellular actions of NSAIDs that contribute to their untoward effects (Rainsford 2004a; Bjarnason et al. 2007). The ratio of inhibition of COX-1 to COX-2 has been considered to relate to likelihood of developing upper GI and possibly renal and other reactions by NSAIDs in relation to their anti-inflammatory activities (Cryer and Feldman 1998; Warner et al. 1999; Vane and Botting 2001). The newer class of highly selective COX-2 inhibitors, the coxibs, were developed in attempts to reduce the risks of serious upper GI and other reactions (Rainsford 2004b).

It has been suggested that one the reasons for the low gastro-ulcerogenicity of ibuprofen may relate to the competition of the COX-1 inactive R(−)-isomer with the active S(+)-enantiomer for the active site of COX-1 so effectively diminishing the potential for inhibition of PG synthesis by the drug (Rainsford 1999b, 2003). Thus, presence of the R(−)-ibuprofen in the ingested drug may mask the interaction of the S(+)-isomer with the active site of COX-1 in the stomach and platelets circulating through the gastric circulation so reducing the inhibitory potential of the latter isomer on production of gastric prostaglandins (Rainsford 1999b). The short plasma elimination half-life of the drug may also be a feature accounting for low risks of upper GI injury form the drug (Henry et al. 1996).

To establish what could be regarded as “clinically-significant” effects of NSAIDs/coxibs on COX-1 and COX-2 activities in humans requires use of either ex vivo whole blood or in vivo blood or tissue sampling techniques that are now well-established (Rainsford et al. 1993; Warner et al. 1999; Brooks et al. 1999; Shah et al. 2001). Using the whole blood assay Blain et al. (2002) compared the effects of ibuprofen, diclofenac and meloxicam on in vitro activities of COX isoenzymes using blood from 24 healthy male volunteers and the ex vivo production of COX-1-derived TxB2 during clotting and COX-2-derived PGE2 upon stimulation with endotoxin after the same volunteers took single and multiple (3 days for ibuprofen and diclofenac and 5 days with meloxicam) doses of 400 mg ibuprofen (Brufen™), 75 mg diclofenac SR (Voltaren™) or 7.5 mg meloxicam (Mobic™). Plasma concentrations of the drugs and in the case of ibuprofen the R/S isomers were determined and used to relate these as free and unbound concentrations to in vitro inhibition profiles.

Ibuprofen taken as a single dose or repeatedly inhibited production of TxB2 by COX-1 by 96 and 90%, respectively. COX-2 production of PGE2 was inhibited by 84 and 76%, respectively after single and multiple doses. The plasma concentrations of R(−) and S(+)-ibuprofen at these times were 116.2 (SD, ±38.7) and 70.4 (±1.28) μM, respectively. Thus, almost complete inhibition of both COX-1 and COX-2 was achieved under in vivo conditions and this was paralleled by the modelling of in vitro inhibition profiles. Near complete COX-2 inhibition in vitro was achieved at free concentrations of the racemate as well as the S(+) enantiomer which almost completely inhibited both COX-1 and COX-2 in vitro. When these data were modelled with the inhibition ex vivo it was evident that although there was wide scatter of about 10% of the latter data most fell within about 80% inhibition of COX-1 and COX-2 observed in vitro. In comparison after oral intake diclofenac inhibited COX-1 by 70% and COX-2 by 95 and 97%. COX-1 inhibition from meloxicam was 30 and 55% and COX-2 was 63 and 83%, respectively after single and repeated doses.

A study relating pharmacokinetics of ibuprofen to analgesic properties from recordings of somatosensory evoked potentials (EP) and subjective pain ratings (PR) following tooth pulp stimulation was performed using a randomized, double-blind, placebo-controlled 6-way crossover study using ibuprofen 400 mg or flurbiprofen 100 mg each taken by six healthy volunteers on two separate occasions compared with placebo (Suri et al. 1997). A comprehensive analysis of the R(−) and S(+) enantiomers of these drugs was also undertaken and the data used to model the PK/PD properties of these drugs. Both EP and PR data correlated to show that peak analgesia was evident at 2.5 h for ibuprofen; the EP data showed prolonged effects of ibuprofen up to 6 h while the PR data were less so but still extended to the same period (Fig. 5). While no data were analyzed in relation to the enantiomers both these showed peak values of R(−) at 1.33 h and S(+) at 2.18 h inferring that the latter isomer is probably the active form in analgesia in accordance with its COX inhibitory properties.
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig5_HTML.gif
Fig. 5

Evoked potentials (EP) and time-dependent development of analgesia to ibuprofen. The peak trough effects are coincident with peak concentrations of R(−) and S(+)-ibuprofen. From Suri et al. 1997; redrawn and reproduced with permission of the publishers of Int J Clin Pharmacol Ther

The data obtained in these studies gives a clear basis for relating the in vivo effects of ibuprofen on prostanoids from COX-1 and COX-2 activities with the anti-inflammatory and analgesic effects of the drug in various clinical models and during therapy. Since the doses of ibuprofen used in these studies (1.2 g day−1) correspond to those used OTC and these doses produce effective analgesic activity in various pain models, notably also inflammatory states such as dental pain involving extraction of molars or throat pain where there is local inflammation it can be concluded that inhibition of COX-2 as well as COX-1 (Martínez et al. 2002) underlies the therapeutic effects of ibuprofen in these conditions (Boureau 1998). It could be argued that the studies of Blain et al. (2002) were in normal volunteers and during inflammatory pain there is likely to be more COX-2 activity. However, the in vitro inhibition profiles modelled against plasma profiles of R(−) and S(+) ibuprofen determined before and after surgery suggest that these concentrations as well as free concentrations of the drug would be sufficient to achieve about 70–80% inhibition.

Thus, this evidence shows that there is a direct relationship between inhibition of the synthesis of inflammatory prostanoids with the dose range of ibuprofen that corresponds to that used in OTC conditions for the relief of minor aches and pain; a central tenet to the therapeutic actions of ibuprofen in the context of the OTC dosage of the drug.

Efficacy

Ibuprofen has been studied extensively in adults and children in a variety of painful and inflammatory conditions where its therapeutic effects are comparable to other NSAIDs and with that of paracetamol; in some studies ibuprofen is slightly superior to that of paracetamol, especially if there is a pronounced inflammatory component to the condition being treated (McQuay and Moore 1998; 2006).

In children, ibuprofen is principally used for control of febrile reactions accompanying infections and minor pain states. Extensive controlled clinical trials and observational studies have been performed in these conditions. Overall, ibuprofen has comparable or greater antipyretic and analgesic effects in children compared with that of paracetamol.

Treatment of pain in osteoarthritis, rheumatoid arthritis and related conditions

There are indications that ibuprofen is used quite extensively in some countries where it is available for OTC use for occasional treatment of osteoarthritis (OA) and other musculoskeletal conditions. Here ibuprofen finds particular value in being more effective than paracetamol and having less GI symptoms than aspirin. Evidence in support of ibuprofen being effective in relieving pain and joint symptoms at OTC dosage in OA has come from various clinical trials; the earlier studies showing effectiveness in this condition were reviewed by Kean et al. (1999).

Among these studies, Bradley et al. (1991) found that 1 month’s treatment with ibuprofen 1,200 mg day−1 (referred to as an analgesic dose) had superior rest and walking pain scores and comparable Health Assessment Questionnaire (HAQ—Stanford University) pain and walking scores with paracetamol (acetaminophen) 4,000 mg day−1 in patients (>70% females; total number of 182 subjects) with OA of the knee. Adverse events were minor symptomatic and comparable with both drugs. The higher prescription-level dose of 2,400 mg days−1 ibuprofen (referred to as an anti-inflammatory dose) resulting in decreased pain and walking scores. A later re-analysis of this study by the same group (Bradley et al. 1992) attempted to resolve the issue of whether the drug effects on joint tenderness and swelling, reflecting synovitis, being affected by “anti-inflammatory” doses of ibuprofen without any indications of any joint symptoms reflecting anti-inflammatory effects. A later re-evaluation of the earlier study by the same group (Bradley et al. 2001) showed that baseline pain could influence the anti-inflammatory/analgesic effects of ibuprofen with better anti-inflammatory response being observed when there was higher levels of baseline pain. This is relevant since there were indications that the original study population (Bradley et al. 1991) comprised patients with mild-moderate joint conditions. Thus, the effectiveness of ibuprofen is greater in patients with more pronounced joint symptoms.

The semantic definition of analgesic effects at OTC dosage of ibuprofen (1,200 mg) being separate from anti-inflammatory effects of this drug (Bradley et al. 1991) is probably not justified since there are several studies in chronic arthritic conditions showing that joint symptoms are significantly improved with 800–1,200 mg day−1 ibuprofen (Kean et al. 1999). Moreover, there is an important study performed by Deodhar et al. (1973) on the effects of ibuprofen 1,200 mg day−1 for 1 week on joint inflammation in RA patients in which they investigated knee inflammation following i.v. injection of radioactive pertechnate (99mTc). The uptake of 99mTc into knee joints was significantly reduced by ibuprofen compared with placebo. A correlation was observed between inflammatory indices of knee functions and 99mTc uptake. This is direct evidence for local anti-inflammatory effects of ibuprofen under conditions where there is effective relief of joint inflammatory symptoms and pain in arthritic diseases. Such relief of joint inflammatory pain is not evident with high doses of paracetamol.

In another short-term study designed to measure pain relief (measured using a 100 mm Visual Analogue Scale, VAS) over the 8-h period following the first and sixth dose (total 1,200 mg day−1) of ibuprofen 200 mg tablets were compared with placebo or ibuprofen + codeine (30 mg) to 29 patients with knee OA (Quiding et al. 1992). Relatively rapid pain relief was evident with ibuprofen (as well as ibuprofen + codeine) and this was sustained throughout the 8-h period of evaluation. These results attest to the specific time-course effects of ibuprofen in relief of pain in OA.

Schiff and Minic (2004) compared the effects of 1,200 mg day−1 ibuprofen with naproxen sodium 660 mg day−1 at an OTC dose (another NSAID that is available OTC in some countries) and placebo for 7 days in 2 multicentre parallel group studies in 440 patients with knee OA. They found that both NSAIDs reduced joint symptoms with naproxen being slightly more effective than ibuprofen. However, ibuprofen was more effective in relief of day pain.

Another large multi-centre study (known as Ibuprofen, Paracetamol Study in Osteoarthritis or IPSO) by Boureau et al. (2004) in 222 patients with OA of the hip (30% patients) or knee (70% patients) ibuprofen 400 mg taken as single or multiple (1,200 mg day−1) doses was compared with paracetamol 1,000 mg single dose or 4,000 mg day−1 for relief of pain and joint symptoms, WOMAC (Western Ontario McMaster University) scores over 2 weeks. Significant reduction was observed in Pain Intensity Scores over the first 6-h period and then was progressively reduced over the 2 week period with ibuprofen compared with that of paracetamol. This study shows that ibuprofen is superior to paracetamol at OTC doses in relief of joint symptoms in both knee and hip OA. This conclusion is supported by a more general Cochrane Review of randomised and placebo-controlled trials in which NSAIDs (including ibuprofen) were superior to paracetamol in achieving reduction in pain, global efficacy assessments and improvements in functional status (Towheed et al. 2006).

Thus, of the available OTC treatments, ibuprofen would appear to have particular advantage for self-treatment of joint pain and symptoms in OA in being superior in efficacy to paracetamol and preferable to aspirin because of a lower incidence of GI symptoms (see later section on adverse reactions).

Outcomes from large-scale clinical trials with coxibs

The studies with the coxibs conducted during the past decade were undertaken with large numbers of patients under modern standards of clinical investigation and with demanding requirements to establish safety in the GI, CV and other organ systems where serious adverse events with the NSAIDs often occur at low frequencies. Ibuprofen was used in a number of these studies as a “bench standard” in recognition of it being accepted as amongst the safest of all NSAIDs that is still widely used in rheumatological and other musculo-skeletal conditions. These studies have afforded a useful and high quality source of data for assessment of the adverse reaction and general safety profile of ibuprofen from studies performed with rigorous quality controls and in a setting where the drug is being critically evaluated against competitors.

The individual adverse reactions in GI, CV and other organ systems are reviewed in detail in later sections. Here, the adverse reaction profiles for ibuprofen and comparator drugs are viewed in a global sense employing outcome measures that are considered good indicators of overall patient and physician acceptability for safety and efficacy. It is important to note that withdrawal of an NSAID from use can reflect occurrence of serous adverse events as well as lack of efficacy.

In an evaluation of the tolerability in adverse events in clinical trials conducted with the objective of assessing celecoxib (Celebrex®, Pfizer) in osteoarthritis and rheumatoid arthritis Moore et al. (2005) used data from the manufacturer’s database (Pfizer) of clinical trials for comparing the occurrence of adverse events, clinical responses and discontinuations in the treatment of arthritis from lack of efficacy or side effects of celecoxib with those from ibuprofen, diclofenac, naproxen, paracetamol and rofecoxib. Although there are limited data available on ibuprofen the data reveals that discontinuation due to adverse events with ibuprofen following 12 or 24 plus weeks of treatment were similar to those from diclofenac or celecoxib when either the number of events or the percentage of discontinuation is compared (Table 8). These data should be evaluated in relationship to the 95% confidence interval ranges which notably overlap. The lack discontinuation due to lack of efficacy over 12 weeks was lowest with rofecoxib and diclofenac, followed by ibuprofen which had a low rate of discontinuation at prescription level of the drug, then celecoxib at doses of 100–400 mg day−1. There was a high rate of discontinuation when the drugs had been taken for 24 plus weeks (Table 8). These results suggest that ibuprofen has a relatively low rate of adverse event discontinuation compared with the other NSAIDs or coxibs and that this is not impacted by lack of efficacy.
Table 8

Rates of discontinuations in treatment of arthritis in the celecoxib manufacturer’s trials because of lack of efficacy, or from adverse events

Duration (weeks)

Treatment

Dose (mg day−1)

Number of events

Total number

Discontinuations, % (95% CI)

Number of events

Total number

Discontinuations % (95% CI)

12

Pla

 

521

1,135

45.9 (43.0–48.8)

70

1,135

6.2 (4.8–7.6)

Ibu

2,400

145

692

21 (18.1–23.9)

52

692

7.5 (5.5–9.5)

Dic

100/150

571

6,094

9.4 (8.6–10.2)

488

6,094

8.0 (7.4–8.6)

Nap

1,000

492

6,166

8.0 (7.4–8.6)

590

6,166

9.6 (8.8–10.4)

Cel

100

128

435

29.4 (25.1–33.7)

28

435

6.4 (4.0–8.8)

Cel

200

1

132

0.8 (0.0–2.4)

13

132

9.8 (4.7–14.9)

Cel

400

374

2,399

15.6 (14.2–17.0)

316

2,399

13.2 (11.8–14.6)

Cel

800

120

4,311

2.8 (2.2–3.4)

338

4,311

7.8 (7.0–8.6)

Rof

25

14

345

4.1 (1.9–6.3)

37

345

10.7 (7.4–14)

24

Ibu

2,400

456

1,985

23.0 (21.2–24.8)

456

1,985

23.0 (21.2–24.8)

Dic

100/150

331

2,325

14.2 (12.8–15.6)

593

2,325

25.5 (23.7–27.3)

Cel

400

26

326

8.0 (5.1–10.9)

34

326

10.4 (7.1–13.7)

Cel

800

691

3,987

17.3 (16.1–18.5)

892

3,987

22.4 (21–23.8)

Pla placebo, Ibu ibuprofen, Dic diclofenac sodium, Cel celecoxib

From Moore et al. (2005)

Similar data available from a large scale randomized trial of the efficacy and tolerability of rofecoxib versus ibuprofen in patients with osteoarthritis by Day et al. (2000) showed that patients that received ibuprofen 2,400 mg day−1 for 6 weeks had rates of discontinuation through adverse events of approximately 12% and through lack of efficacy of approximately 3% compared with those of rofecoxib where the discontinuation from adverse events were approximately half these values from ibuprofen but lack of efficacy was comparable. This is an important observation since it has often been argued that the lower rates of ADRs and toxicity of ibuprofen (including that in the GI tract) may be a consequence of the drug being less potent or that it may have differing patterns of prescribing compared with that of other NSAIDs. The evidence is, however, that the anti-inflammatory and analgesic effects of ibuprofen are comparable to that of other NSAIDs when given at recommended prescription levels (Kean et al. 1999). It its true that drugs such as diclofenac and ketoprofen are more potent prostaglandin synthesis inhibitors than ibuprofen and that the selective COX-2 activities of the coxibs such as celecoxib and etoricoxib may reflect greater efficacy of these drugs. However, it is more likely that the longer plasma half-lives of drugs such as naproxen and celecoxib may contribute to these drugs having more sustained analgesic and anti-inflammatory activity compared with that of ibuprofen thus indicating that it may be a question of the duration in circulation of these drugs that accounts for any differences in their therapeutic effects. There is little available evidence to support these concepts and therefore they can only be considered hypothetical.

Another study with a coxib, lumiracoxib 400 mg (Prexige®, Novartis) was compared with ibuprofen 2,400 mg and naproxen 1,000 mg taken for 52 weeks in 18,325 patients randomized for the treatment of osteoarthritis (Farkouh et al. 2004; Schnitzer et al. 2004). There were two major studies performed one addressing the CV events (Farkouh et al. 2004) and the other GI safety (Schnitzer et al. 2004), with each of these having two sub-studies, one a comparison of lumiracoxib with ibuprofen and the other comparing the former with naproxen. Only the data from the ibuprofen sub-study is reviewed here although it is interesting that aside from CV events ibuprofen had lower or comparable safety with naproxen in the occurrence of other ADRs.

Here, the numbers of discontinuations and the reasons for withdrawal are considered (Table 9). It is apparent from this data that there were similar rates of discontinuation from ibuprofen compared with that from lumiracoxib; the same was also evident with naproxen. The losses were mainly due to adverse events, these being relatively low compared with what would be expected in a study lasting 1 year. Likewise, the losses due to unsatisfactory therapeutic effects are low and comparable with one another.
Table 9

Discontinuations in the TARGET study: comparisons of Ibuprofen with Lumiracoxib. Redrawn and produced with permission of Lancet

 

Lumiracoxib vs ibuprofen sub-study

Lumiracoxib (n = 4,399)

Ibuprofen (n = 4,415)

Discontinued

1,751 (40%)

1,941 (44%)

Reason for discontinuation

Adverse events

699 (16%)

789 (18%)

Abnormal laboratory values

42 (1%)

33 (1%)

Abnormal test procedure results

11 (<1%)

13 (<1%)

Unsatisfactory therapeutic effect

393 (9%)

429 (10%)

Patient’s condition no longer requires study drug

11 (<1%)

16 (<1%)

Protocol violation

187 (4%)

192 (4%)

Patient withdraw consent

363 (8%)

419 (9%)

Administrative problems

21 (<1%)

22 (1%)

Lost to follow-up

15 (<1%)

17 (<1%)

Death

9 (<1%)

11 (<1%)

From Schnitzer et al. (2004)

These studies from large scale clinical trails attest to the comparable rates for withdrawals from trials with ibuprofen and the coxibs. They show that the newer coxibs are neither more effective nor less likely to produce adverse reactions leading to withdrawals from studies compared with that of ibuprofen.

Relief of pain, inflammation and fever

It is well established that ibuprofen at both OTC and prescription level dosages is effective in controlling pain and inflammation in a variety of inflammatory and painful conditions. Among these are rheumatic and other musculo-skeletal conditions (Boardman et al. 1967; Owen-Smith and Burry 1972; Altman 1984; Dieppe et al. 1989; Haase and Fischer 1991; Bliddel et al. 2000; Singer et al. 2000; Buchanan and Kean 2002), dental pain and surgery, dysmenorrhoea, upper respiratory tract conditions (colds, influenza), headaches, accidental sports injuries and surgical conditions (Nørholt et al. 1998; Kean et al. 1999; Dionne and Cooper 1999a, b; Rainsford 1999c; Hersh et al. 2000a, b; Steen Law et al. 2000; Doyle et al. 2002; Beaver 2003; Dalton and Schweinle 2006; Sachs 2005; McQuay and Moore 2006; Eccles 2006; Verhagen et al. 2006; Huber and Terezhalmi 2006). It is not the purpose here to review the evidence in extenso about the clinical efficacy in various painful states since this is well established from over 40 years of research and applications of ibuprofen in the treatment of these conditions (Kean et al. 1999; Rainsford 1999c). Ibuprofen is very effective in controlling fever both in adults and in children (Hersh et al. 2000b; Eccles 2006) and is a preferred drug among the OTC analgesics for treatment of tension-type headache (Verhagen et al. 2006). It has wide applications in the treatment of viral respiratory infections where there is an appreciable inflammatory component (Winter and Mygynd 2003). There are, however, some key points about the efficacy of ibuprofen which need to be emphasized in the context of the OTC use of this drug and comparisons with other analgesics, as follows:
  1. 1.

    Ibuprofen has rapid onset of analgesia and this is maintained in parallel with the plasma elimination half life of ibuprofen which for both the active (S+) and inactive (R−) enantiomers is approximately 2 h (Graham and Williams 2004); the analgesia extending to approximately 6 h as evidenced from a number of analgesia models (e.g. third molar dental extraction pain model; Dionne and Cooper 1999).

    As mentioned previously in acute pain there are alterations in the pharmacokinetics of ibuprofen resulting in decreased serum levels of the enantiomers after dental surgery (Jamali and Kunz-Dober 1999). Gender differences have been observed in response to acute pain in the dental pain model with ibuprofen being more effective in men than in women (Walker and Carmody 1998). Other studies using a similar third molar extraction dental pain model have not revealed any gender differences in response to analgesia with ibuprofen (Averbuch and Katzper 2000). Aside from these factors it appears therefore that variability in response to analgesia with ibuprofen may relate to acute pain and altering the pharmacokinetics of the drug and possibly gender differences.

     
  2. 2.

    Ibuprofen has been found to be effective in chronic arthritic pain with associated improvement in joint inflammatory symptoms even at low (800–1,200 mg day−1) doses (Kean et al. 1999). Under these conditions paracetamol is less effective. Indeed several studies have shown that paracetamol has little or no effects in controlling chronic inflammatory pain in rheumatic diseases at the recommended OTC dosages of 1,000 mg day−1 (Kean et al. 1999).

     
  3. 3.

    There is unequivocal evidence for a dose–response effect with ibuprofen in acute pain conditions (McQuay and Moore 2006; Wan Po 2006). Thus, a meta-analysis of McQuay and Moore (2006) only showed that 400 mg day−1 ibuprofen was superior to 200 mg. More extensive analysis of studies by Schou et al. (1998), showed that there was a much greater range of dose–response that extended from 100 to 400 mg. Using the number needed to treat (NNT) analysis for ibuprofen was in the range of 6–23 for the high to low dose comparisons compared with that of paracetamol 1,000–500 mg which was 6–20; the inference being that there is no difference in the NNT between these two treatments. However, a consensus view is that in certain inflammatory pain conditions (e.g. dental surgery) ibuprofen is superior at its recommended OTC dosage of 200–400 mg per single treatment compared with that of paracetamol 500–1,000 mg (Dionne and Cooper 1999; Hargreaves and Keiser 2002; McQuay and Moore 1998).

     
  4. 4.

    In comparison with other NSAIDs, including the newer coxibs, ibuprofen has been shown in a variety of studies to be at least as effective as these drugs with the possible exception that longer half-life drugs such as rofecoxib exhibit a longer duration of action (Dionne and Cooper 1999; Huber and Terezhalmi 2006; Hargreaves and Keiser 2002; Edwards et al. 2004) and as well, ibuprofen as been shown to be effective in a variety of acute pain conditions (Sachs 2005).

     

A key pharmacokinetic feature about the analgesic actions of ibuprofen is that the drug has the ability to penetrate the CNS and is present in free (i.e. non-protein bound) concentrations in the CSF (Brocks and Jamali 1999; Graham and Williams 2004). As well ibuprofen accumulates and is retained in inflamed joints of arthritic patients (Brocks and Jamali 1999; Graham and Williams 2004). Thus the drug is present in sites where analgesic and anti-inflammatory effects are required.

Recent reviews of the choice of analgesics for pain management have highlighted that although paracetamol may be useful and perhaps a choice initially for pain control that the safest and most effective of all NSAIDs is ibuprofen at doses of 400 mg for acute non-specific pain (Sachs 2005) and is highly effective in tension-type headache (Verhagen 2006). A particularly challenging opinion provided by a recent analytical review by Arora et al. (2007) suggested that oral ibuprofen is as effective in analgesia as parenteral ketolorac, a drug which is used in postoperative surgical pain control as well as in acute traumatic musculo-skeletal pain conditions. Since ketolorac is amongst the NSAIDs with the highest risks for showing upper gastrointestinal bleeding and ulcers, it appears that substitution of oral ibuprofen in acute surgery and traumatic conditions may represent a valid alternative with much lower risk than parenteral ketolorac.

Acute fever in children

General aspects concerning treatment of fever in children

Ibuprofen is widely used in the treatment of fever and in the treatment of this condition care-providers can have a considerable involvement. Likewise, paediatricians and general physicians as well as pharmacists have a role in the application or prescription of ibuprofen for the treatment of fever. While ibuprofen has second place in the treatment of fever over the past 3–4 decades, paracetamol has found wide popular application and is considered to be a safe and effective treatment for most febrile conditions. However, there are indications from some clinical trials that ibuprofen, especially in the long-term, may be more effective than paracetamol (van den Anker 2007). There are also indications that alternating treatment with paracetamol and then ibuprofen or combinations of these two is becoming increasingly popular especially amongst paediatricians and those involved in the treatment of children under emergency or outpatient conditions (Sarrell et al. 2006). The application of combinations or alternating treatments with paracetamol and ibuprofen is highly controversial and is regarded by most as having potential risks. Indeed this author believes that there is a major issue concerning potential toxicity in certain organs for example in the liver and kidney. Paediatric patients with very high febrile states that lead to cytokine activation of liver reactions, may be at risk following paracetamol administration. Monotherapy is generally preferred and ibuprofen has a key place in treatment of fever in infants and children.

Despite these apparent benefits the application of antipyretic agents to treat fever in infants and children has not been without its critics, furthermore parents and care-providers have been considered to have numerous misconceptions about what fever is and how it should be treated (Crocetti et al. 2001; Stagnara et al. 2005). One leading US physician, Barton-Schmitt MD, some 20 years ago found that parents had many misconceptions about what fever really is in terms of temperature and he coined the term “Fever phobia” (Crocetti et al. 2001). In a survey of 340 care-providers in two urban based paediatric clinics in the Baltimore region, Maryland USA, considerable variation was evident in potential harm that fever cause their children, or even what temperature range actually constitutes fever and in the application of sponging and other treatments to control fever.

In an editorial in the British Medical Journal, Hay et al. (2006) reviewed some of the recent studies on single and combination antipyretic therapies and highlighted that the safety of combinations is limited. These authors noted the occurrence of renal failure or renal tubular necrosis from ibuprofen and the potential for nephrotoxic metabolite formation from paracetamol (quinine–imine paracetamol) in producing both nephrotoxicity and hepatic reactions. They also pointed out that the definition of clinically useful difference in temperature after treatment is still debatable. To achieve better understanding continuous thermometry should be employed. Hay et al. (2009) have recently reported cost benefit analysis and a trial in a UK National Health System (NHS) setting on the application of ibuprofen alone or in combination with paracetamol for treatment of fever in children. On the basis of this investigation Hay et al. (2009) recommended in unwell patients with fever that ibuprofen be used first, but that the combination may be employed where longer duration of control of fever is required but the risks of toxicity should be considered. Both ibuprofen alone or the combination were superior to paracetamol.

Dlugosz et al. (2006) have reviewed the appropriate use of non-prescription antipyretics in paediatric patients. They referred to the ongoing debate about whether and when to treat fever, but clinicians agree that antipyretic therapy is important for febrile children who have (a) chronic cardio-pulmonary disease, metabolic disorders or neurological conditions, and (b) are at risk for febrile seizures. They point to the lack of guidelines on the use of antipyretic agents in other categories of fever in children. Thus, patient comfort is cited most often as the deciding factor. Moreover, there is little support for administering antipyretic agents when the temperature is less than 38.3°C (101°F) unless the child is uncomfortable. None-the-less they regard paracetamol and ibuprofen as effective agents for reducing fever and this is supported by evidence from meta-analyses and other studies. They also point to risks of paracetamol hepatotoxicity especially in children with diabetes, those with concomitant viral infections, patients with a family history of hepatotoxic reactions, obese children and chronically malnourished individuals. Dlugosz et al. (2006) also emphasized the precise dosing of paediatric patients with either ibuprofen or paracetamol (based on recommendations of the American Academy of Pediatrics) and in the case of a patient less than 6 months with a temperature ≥37.9°C (100.2°F) recommend immediate consultation of a physician or paediatrician.

The application of ibuprofen and other antipyretics to prevent the development of febrile seizures is now a well-established treatment for this condition (van Stuijvenberg et al. 1998a, b, 1999).

The question of precise dosage of antipyretics for treatment of fever and pain has been addressed by a number of experts and professional organisations. Among these the Royal College of Paediatrics and Child Health with the Neonatal and Paediatric Pharmacists Group in their monograph “Medicines for Children” (2003) (see http://www.rcpch.ac.uk/Publications/bnfc; accessed 16 Nov 2009) recommend for pyrexia and mild to moderate pain that ibuprofen is given by the oral route and that the dosage should be in relation to body weight (5 mg kg−1) 3–4 times daily when treating infants, and children from 1 month to 12 years of age. Dosage by age is recommended above 1 year and for 1–2 years 50 mg, 3–7 years 100 mg, and 8–12 years 200 mg of ibuprofen. In 12–18 year age group 200–600 mg ibuprofen is recommended 3–4 times daily. In juvenile rheumatoid arthritis or juvenile arthritis application of ibuprofen is recommended at the dose of 10 mg kg−1 for the 1 month to 18-year-old group 3–4 times daily or up to 6 times daily in systemic juvenile arthritis.

In Martindale (2002) “The Complete Drug Reference” ibuprofen is not recommended for children below 7 kg bodyweight in the same way as with the previous authors dosage on a bodyweight basis but is recommended in the range of 20–30 mg kg−1 day−1 in divided doses or alternatively in the 6–12 m age group 150 mg day−1, 1–2 year 150–200 mg day−1, 3–7 year 300–400 mg day−1 and 8–12 year of age 600–800 mg day−1. These authorities clearly differ in the recommendations in for treating fever and pain in children on a dosage basis. Arguably, however, dose recommendations are probably rather similar and it is a question of the application of information that is given to the care giver.

The UK National Institute for Health and Clinical Excellence (NIHCE) has prepared recommendations for the assessment and management of children younger than 5 years in their report “Feverish Illness in Children” (2007) (see http://www.guidance.nice.org.uk/CG47/NiceGuidance/pdf/English, accessed 16 Nov 2009). They have prepared two reference guides “Understanding NICE Guidance” one of which is a quick reference guide (N1247) and the other is “Understanding NICE Guidance” (N1248). In the NICE clinical guideline No. 47 emphasis is given on the detection of fever and the clinical assessment of the child with fever as well as the relative roles of the non-paediatric practitioner and paediatric specialist.

Surprisingly the NICE recommendations state that antipyretics do not prevent febrile convulsions and should not be used specifically for this purpose. This is in contraindication to the published evidence which show that antipyretic agents do have a role in the control of febrile convulsions. Their recommendations also give a considerable number of clinical diagnostic indices for fever of various origins. Some of these recommendations are complex in themselves. Among them “A Traffic Light System” for identifying risks of serious illness involving colour coding of green for low risk, amber for intermediate risk and red for high risk with appropriate diagnostic and investigative procedures for identifying the origin of fevers.

The NICE recommendation for antipyretic interventions state that tepid sponging is not recommended for the treatment of fever. This is in contrast with recommendations of other authorities. On the question of the application of antipyretic agents these should be considered in children with fever who appear distressed or unwell. Antipyretic agents should not routinely be used with the sole aim for reducing temperature in children with fever who are otherwise well. The views and wishes of parents and care givers should be taken into consideration; this would in any physician’s eyes be regarded as a statement of the obvious!

NICE recommendations are that either paracetamol or ibuprofen can be used to reduce temperature in children with fever but that they should not be given at the same time or alternatively. The only case for alternative drug treatment would be considered for a child that does not respond to the first agent.

The guideline development group (GDG) has made a number of recommendations for research based on the review of evidence to improve NICE guidance for patient care in the future for example predictive values of heart rate, remote assessment and a number of issues concerning diagnosis. They also recommend investigation of the application of antipyretics in primary and secondary settings in relationship to the degree of illness.

Evidence for antipyretic efficacy of ibuprofen and comparisons with other antipyretic agents

The original developer of ibuprofen (Boots Ltd., Nottingham) undertook a number of clinical trials in children with various febrile conditions in various countries during the 1980s which involved investigating the effects of ibuprofen in comparison with aspirin or paracetamol for the treatment of fever in children. These studies were done under conditions that were regarded as being quite reasonable for the time but would now probably not be to the high standards of GCP today. None-the-less these studies have given an important basis for showing the effectiveness of ibuprofen as well as giving some information on safety. It is not proposed to review these early studies at length although some key points will be considered in this section. More detailed consideration will be provided to more recent published information which is giving some insight into the relative efficacy of ibuprofen in various febrile conditions as well as its safety.

Among the early studies was that by Kandoth et al. (1983) of the KEM Hospital, Bombay, India. This study was an open-label comparative investigation of ibuprofen suspension with that of aspirin in a cross-over design in 28 children suffering from fever due to upper respiratory tract infections or other causes and whose rectal temperature was about 38°C and having been admitted to the hospital. In this small study, rectal temperatures of those patients given a daily dose of ibuprofen 20 mg kg−1 or aspirin 45 mg kg−1 bodyweight were found to be approximately equivalent with the reduction in mean body temperature being observed at 1 h with the maximum fall at 3 h after administration. There did not appear to be any record of adverse reactions.

Another study by Amdekar and Desai (1985), the antipyretic activity of ibuprofen was compared with that of paracetamol in 25 children suffering from fever due to upper respiratory tract infection or systemic viral infections. There was a difference in the initial temperatures of patients that were treated for upper respiratory tract infection in that the mean initial temperature was 39.9°C in the ibuprofen group and 40.81°C in the paracetamol group. Despite this difference both ibuprofen and paracetamol produced statistically significant reduction in rectal temperatures following administration of 7 mg kg−1 of bodyweight of ibuprofen or 8 mg kg−1 bodyweight of paracetamol; the application of these been given in a random order. Initial reduction in temperatures of patients with upper respiratory tract infections incurred at 0.5–1 h with the maximum at 4 h after administration of both drugs. The level of antipyretic activity was evident up to 8 h with patients having temperatures in the range of 37.5°C. In the group of children with fever due to viral infections both the mean temperatures were comparable (40.51–40.75) and similar results were observed as in the patients with upper respiratory tract infections. The exception is that by 8 h the temperatures had risen to 38.34–38.77°C which is somewhat higher than that observed in the patients with upper respiratory tract infections and probably reflects ongoing viral activities.

A single blind, parallel group investigation comparing the antipyretic properties of ibuprofen syrup versus aspirin syrup in 78 febrile children aged 6 months to 10 years was undertaken in two centres in Belgium by Heremans and co-workers (1988). Using doses of ibuprofen syrup (6 mg kg−1 body weight) or aspirin 10 mg kg−1 bodyweight, significant reduction in rectal temperature were observed with both treatments; there being no statistically significance between the two. These patients had a greater variety of clinical history although most were being treated for upper respiratory tract infections in some cases with antibiotics being co-administered. Significant reductions in temperatures were observed by 0.5–1 h with both treatments and maximum reduction in rectal temperature being observed with both drugs at 4 h and being maintained 6 h after administration.

A summary of more recent data from various studies reviewed by Autret-Leca (2003) is shown in Table 10. These studies have been performed with modern methodologies and in some cases under GCP conditions. Dose-ranging studies (Table 11) show that ibuprofen is effective in the recommended dosages that are over a wide range. The results show that ibuprofen is equal to or in some cases slightly more effective than paracetamol in relief of febrile symptoms in a variety of age groups in children.
Table 10

Comparative studies of ibuprofen and paracetamol in the treatment of acute fever

 

Dose frequency

Age (years)

No of children

Dose of ibuprofen (mg kg−1)

Dose of paracetamol (mg kg−1)

Outcome

Sidler et al. (1990)

Multiple

1.25–13

90

7 or 10

10

Ibu 7 > Para

Ibu 10 > Para

Wilson et al. (1991)

Single

0.25–12

178

5 or 10

12.5

Ibu 10 > Para

Autret et al. (1994)

Multiple 3 days

0.5–5

154

7.5

10

Ibu = Para

Van Esch et al. (1995)

Multiple 3 days

0.25–4

70

7.5

10

Ibu > Para

Vauzelle-Kerroëdan et al. (1997)

Single

4 ± 0.6

116

10

10

Ibu = Para

Autret et al. (1997)

Single

0.5–2

351

7.5

10

Ibu > Para

From Autret-Leca (2003)

Table 11

Dose-ranging studies of ibuprofen for the treatment of acute fever

Author (year)

No of children

Age range (years)

Dose (mg kg−1)

Efficacy (dose)

Kotob et al. (1985)

44

2–12

5, 7.5

7.5 > 5

Walson et al. (1989)

127

2–11

5, 10

10 > 5

Sidler et al. (1990)

90

0.4–13

7, 10

10 > 7

Wilson et al. (1991)

178

0.25–12

5, 10

10 > 5

Marriott et al. (1991)

93

Mean 2.5

0.625, 1.25, 2.5, 5

Kauffman and Nelson (1992)

37

2–12

7.5, 10

7.5 = 10

Nahata et al. (1992)

56

0.4–12

5, 10

10 > 5

Walson et al. (1992)

64

0.5–11

2.5, 5, 10

From Autret-Leca (2003)

Pain relief in children

Acute painful or inflammatory conditions

Ibuprofen has been successfully employed in a variety of acute pain models in children (Autret-Leca 2003; Perrott et al. 2004; Table 12). These include pain from post-dental extraction (Moore et al. 1985), tonsillo-pharyngitis (Bertin et al. 1991), otitis media (Bertin et al. 1996), migraine (Hämäläinen et al. 1997; Annequin and Tourniaire 2005; Géraud et al. 2004; Cuvellier et al. 2005), orthodontic separator placement or orthodontic pain (Bernhardt et al. 2001; Bird et al. 2007; Bradley et al. 2007) in which ibuprofen has equivalent or greater efficacy than paracetamol (Table 12). However, a meta-analysis by Perrott et al. (2004) concluded that ibuprofen was superior as an anti-pyretic compared with paracetamol.
Table 12

Comparative studies with ibuprofen in the treatment of acute pain

Indication

No of patients

Age (years)

Dose (mg kg−1)

Results

Dental pain

45

5–12

Ibuprofen 200 mg (6.5 mg kg−1)

Paracetamol 300 mg (10 mg kg−1)

Ibu = Para

Tonsillopharyngitis

231

6–12

Ibuprofen 10

Paracetamol 10

Ibu = Para

Migraine

88

4–16

Ibuprofen 10

Paracetamol 15

Ibu > Para

Juvenile idiopathic arthritis

86

2–15

Ibuprofen 30–40 day−1

Aspirin 60–80 day−1

Ibu = Asp

From Autret-Leca (2003)

Lewis et al. (2004) in a report of the American Academy of Neurology Quality Standards Sub-Committee and the Practice Committee of the Child Neurology Society reviewed the pharmacological treatment of childhood migraine. They concluded that of the treatments available in >6 year olds, ibuprofen 7.5 mg kg−1 (4–16 year or 6–12 year, respectively) is effective and paracetamol 15 mg kg−1 is probably effective for the treatment of acute migraine with both being recommended. The evidence for ibuprofen being most effective being from two double-blind, placebo, Class 1 trials, and was considered a Level A treatment above that of paracetamol (Level B). Triptans are also recommended (Class A) especially in their role as preventative therapies. Ibuprofen for migraine is also recommended in children ≤6 months by the “French Recommendations for Clinical Practice; Diagnosis and Therapy of Migraine” (Géraud et al. 2004). A systematic review of medication trials in children by Damen et al. (2005) considered the evidence showed that ibuprofen and paracetamol with or without triptans was effective in treating childhood migraine.

Ibuprofen has also been found to have beneficial effects in children with musculo-skeletal conditions and post-operative pain (Kokki et al. 1994), transient synovitis of the hip (Kermond et al. 2002), ankle sprains (Dalton and Schweinle 2006) and musculo-skeletal trauma (Clark et al. 2007).

Juvenile idiopathic (chronic) or juvenile rheumatoid arthritis

This condition presents with a varying spectrum of clinical manifestations that include differing joint involvement: including pauciarticular, ≤4 joints; polyarticular ≤5 joints with juvenile rheumatoid arthritis (JRA, Still’s Disease) as a subgroup resembling the adult disease (Dieppe et al. 1985; Klippel 1997). Pyrexia is common (50%) while lymphadenopathy, splenomegaly, pericarditis and rashes may occur (Dieppe et al. 1985). High doses of the NSAIDs, especially aspirin and other salicylates, have been widely used in the treatment of juvenile arthritis and more recently the coxibs (Ansell 1983; Hollingworth 1993; Klippel 1997; Eustace and Hare 2007).

Amongst the earlier studies on the effects of ibuprofen in JRA was that by Ansell (1973). She undertook an open-label investigation in eight patients (aged 7–14 years; 5 female, 3 male) most of them were treated because they were unable to tolerate aspirin and had a prior history of dyspepsia (5) or GI bleeding (2) or in one case where there was poor control. These patients received various doses of ibuprofen (13–32 mg kg−1). Initially they received 200–300 mg day−1 in those with body weight of 20–30 kg and 400 mg for those >30 kg. Later all but one received 600 mg day−1 and one 1,200 mg day−1 for what appears to have been long periods of time (12–24 months). Satisfactory control of pain and stiffness were observed in six of eight cases, although in two of these the dose had to be increased before this was achieved. Occult blood which had been observed in those patients who were on aspirin became negative with ibuprofen. In six patients, liver function tests were performed and none showed increased SGOT, SGPT or alkaline phosphatise and some showed decrease in these values. This is important since plasma/serum levels of elevated liver enzymes have been frequently observed in patients with JRA that have received aspirin and ADR’s in the GI tract and other systems are frequently observed with the salicylates and other NSAIDs in these conditions (Hollingworth 1993; Kean et al. 2004).

Giannini et al. (1990) undertook two studies—one being a multi-centre, randomized, double-blind study in 92 children (76 girls, 16 boys) mean age 7.7 years [range 1.8–15.1 years], mean body weight 26.4 (range 11.5–58.7) kg with JRA. Of these 45 received ibuprofen suspension 30 mg kg−1 day−1 and 47 aspirin 200 mg tablets or 300 mg caplets according to body weight (60 mg kg−1 day−1) for 12 weeks. This was followed by an open-label study in 84 patients (aged 1–15 years, mean 5.3 years; av. body weight 19.9 kg [range 10.0–58.0 kg]). Ten patients failed to complete the double-blind study, nine of whom had received aspirin and ibuprofen; while a further six patients were excluded from the aspirin group due to variation in diagnosis or disease condition. All the patients on ibuprofen showed reduction in all five joint parameters, while those that received aspirin showed significantly and clinically fewer reductions in joint inflammation and pain on motion although the reduction in morning stiffness was the same in both groups.

In the open-label study, 3 dropped out and 16 out of 84 failed to complete the 24 weeks of the study. Time-dependent improvement in overall efficacy scores was observed in all 65 patients that completed the study who received 30–50 mg kg−1 day−1 ibuprofen. One or more ADR’s were observed in 55% patients that were classified as possibly, probably or definitely related to the drug. Upper GI disturbances were recorded in 31 and 27% in the lower tract, with dose-related effects on the former group. Of these three patients had GI bleeding which resolved after discontinuing the drug. Increased serum alkaline phosphatase and bilirubin occurred in two patients that had 40 mg kg−1 day−1 ibuprofen.

Steans et al. (1990) examined the safety, efficacy and acceptability of 10 (initially)–40 (maximum) mg  kg day−1 ibuprofen syrup in 46 children with JRA (aged 18 months–13 years; mean 6.8 years) in a multicentre, open-label study that extended on average for 8 months (range 8 weeks–2+ years). Six patients failed to complete the study, two of which had suspected side-effects. Assessments of active joints and disease activity at monthly intervals over the first 3 months showed that statistically significant reduction in numbers of swollen and/or tender joint at ≤2 months of therapy which progressed to 6 months, while the physician’s VAS was reduced by 1 month and showed significant improvement thereafter which was sustained at 4–6 months. Side effects included gastritis (1 patient), abdominal pain (1 patient), and taste disturbance and nausea (1 patient). Of the 39 children that completed the trial, 28 showed improvement on therapy, 7 were worse and 4 remained unchanged.

Effects of suppositories

A study performed by Stagnara et al. (2005) in Lyon, France was an open-label, multi-centre, non-comparative study designed to evaluate the antipyretic effects and acceptability of 60 or 125 mg ibuprofen suppositories over a 5 day period. The study was performed (in a planned number of 80 patients but involved 45 patients who were analyzed) aged 3 months to 6 years. Assessments were undertaken by both parents and practitioners and involved categorical scale determinations of acceptability and temperature measurements. Temperatures declined within 1/2 h after administration of the suppositories and were lowered for up to 8 h afterwards. Similar reductions in temperatures were evident in 2 year olds compared with 6 year olds showing there was no age dependence in antipyretic effects of the drug. The efficacy of drug treatment was found to be favourable with 88.4% of parents and 86% of physicians reporting that the treatment was “Efficacious” or “Very Efficacious”. The treatment was judged to be “Well Adapted” or “Perfectly Adapted” in the clinical situation by 86% of the investigators. Eight patients had adverse effects which were non-serious. Rectal examinations by the investigators showed that there were no local reactions but parents reported abnormalities in four patients that were minor in which redness was observed but this did not lead to stoppage of treatment.

General safety profile

There were indications of a favourable safety profile for ibuprofen from the post-marketing data during the 15 year period after approval in the USA (Royer et al. 1984). This information was important in decisions made by the FDA in granting approval for OTC use in the USA (Rainsford 1999c).

Since then the safety profile of ibuprofen has been compared with a range of other new and established NSAIDs on the basis of being a recognized “bench mark” for safety and efficacy comparisons (Kean et al. 1999). In particular, ibuprofen has been used as a comparator drug in several trials with the newer coxib class of NSAIDs; an aspect that will be considered later in this section and subsequent sections on the adverse events and toxicity in individual organ systems. Data from epidemiological sources and from a large number of controlled clinical trials performed in comparison with the newer coxib class of NSAIDs has shown that ibuprofen is amongst the tNSAIDs with the lowest risks of serious GI events (i.e. peptic ulcer bleeds, PUBs etc.) (Rainsford et al. 2008). In comparison with coxibs ibuprofen has shown slightly higher risks when given short term (1–3 months) yet after long-term (6+ months) ibuprofen has comparable and low risks of serious GI events. Symptomatic GI events (dyspepsia, nausea, heartburn, diarrhoea, constipation, vomiting) from ibuprofen are, in general, comparable with those from the coxibs. Thus, ibuprofen can be considered to have consistently shown relatively low risks of serious GI events in the recent extensive studies in large scale controlled clinical trials.

Assessments in this section with supporting data and information show (a) the overall safety profile of ibuprofen at the current recommended prescription dosage 1,800–2,400 mg day−1 for the short and long-term treatment of acute and chronic moderate to severe inflammatory pain conditions including rheumatoid- and osteoarthritis, spondylo-arthropathies and other rheumatic conditions; (b) the safety of ibuprofen ≤1,200 mg day−1 for a maximum dosage period of 7–14 days which in a number of countries worldwide is sold as a non-prescription over the counter sale (OTC) analgesic for the relief of mild to moderate painful conditions many of which have a moderate acute inflammatory component; (c) the efficacy and therapeutic activities of ibuprofen principally at OTC dosage; (d) assessment of the risks/benefits of ibuprofen compared with other analgesics (paracetamol) and OTC NSAIDs (ketoprofen, naproxen) that are sold as non-prescription (OTC) analgesics in some countries; and (e) a review of the recommendations and guidelines for the use and safety issues derived from the countries where ibuprofen is a drug sold in considerable quantities for OTC use or General Sales List (GSL) (e.g. USA, UK, Canada, Australia and other Commonwealth countries) with EMEA recommendations and requirements and guidelines on the safety of NSAIDs and analgesics. The conclusion from this assessment of the safety and benefits of ibuprofen can be summarized thus:
  • Ibuprofen at OTC doses has a low possibilities of causing serious GI events, and little prospect of developing renal and associated CV events.

  • Ibuprofen OTC does not pose a risk for developing liver injury especially the irreversible liver damage observed with paracetamol and the occasional liver reactions from aspirin.

  • The pharmacokinetic properties of ibuprofen especially short plasma half-life of elimination, lack of development of pathologically-related metabolites (e.g. covalent modification of liver proteins by the quinine-imine metabolite of paracetamol or irreversible acetylation of biomolecules by aspirin) support the view that these pharmacokinetic and notably metabolic effects of ibuprofen favour its low toxic potential.

  • Moderate inhibition of COX-1 and COX-2 combined with low retention as shown by the mean residence time of the drug in the body may account for the low GI, CV and renal risks from ibuprofen, especially at OTC doses.

There are clearly low risks from OTC ibuprofen based on the safety profile of the drug and the experience in regulatory domains worldwide. The benefits to the public are also clear in providing superior pain-relief in a variety of acute and chronic inflammatory pain conditions with lower risks of toxicity compared with that of the other two most commonly used analgesics, aspirin and paracetamol.

ADRs and safety in prescription level doses in adults

The overall pattern of adverse events from ibuprofen at the prescription dosage-level has been principally derived from studies in adult populations principally those with rheumatic and other acute and chronic painful conditions (Rainsford 1999c). These probably conform to that of all NSAIDs (Rainsford 2001). The diagrammatic representation shown in Fig. 6 indicates the spectrum of adverse reactions that is in general observed with the NSAIDs.
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig6_HTML.gif
Fig. 6

Patterns of adverse reactions from the NSAIDs (Bjarnason et al. 2005). The occurrence and severity of these reactions differs with each drug. Reproduced with permission of Birkhäuser Verlag

This has given rise to the concept that there are NSAID adverse events that can be considered to be class-related (i.e. common to all NSAIDs). Within this concept it is clear that NSAIDs vary considerably in their frequency or occurrence of individual side effects. Some of these are mechanism-related that is to say related to the effects on prostaglandin production via COX-1 inhibition for example in the GI tract and kidneys. However, it has been argued that in relationship to some of these effects that there are important interactions between inhibition of COX-1 and COX-2, nitric oxide synthase and physico-chemical factors which are of significance in the development of these effects (Bjarnason et al. 2007) so that they all can not be considered to be mechanism-related.

Epidemiological studies

A number of studies have been performed examining the relative safety and adverse events attributed to ibuprofen compared with other NSAIDs. Many of these studies have involved examination of the occurrence of adverse events e.g. in specific system organ classes (SOC) or individual reactions in organ systems (e.g. gastro-intestinal ulcers and bleeding). These studies are reviewed in subsequent sections of this review. There are relatively few studies where overall “toxicity” of NSAIDs has been determined under conditions for meeting adequate standards for such epidemiological investigations (e.g. sufficient numbers of study subjects or validity of databases).

Among the more reliable of these was a study by Freis et al. (1991) who performed an investigation of the relative toxicity of a range of NSAIDs used in the treatment of rheumatoid arthritis in the USA with data recorded in the Arthritis, Rheumatism, and Aging Medical Information System (ARAMIS). It should be noted that this study was undertaken in what can be described as the “pre-coxib” era, i.e. before the introduction of coxibs during late 1999. This has significance since the coxibs had an appreciable if variable share of the NSAID market world-wide and thus influenced the overall patterns of use of the NSAIDs in rheumatic and other musculo-skeletal conditions. A summary of the Standardized Toxicity Index from the 11 most-frequently prescribed NSAIDs including ibuprofen in RA patients adjusted for weightings, demographic factors etc., as part of a sensitivity analyses, is shown in Table 13.
Table 13

Toxicity indices of 11 most-frequently prescribed NSAIDs, including ibuprofen in patients with rheumatoid arthritis in data derived from five centres in USA and Canada analyzed from the ARAMIS database (Freis et al. 1991)

Drug

All patients

Drug starts only

No. of courses

Standardized toxicity index score, ±SEM (rank)

No. of courses

Standardized toxicity index score, ±SEM (rank)

Aspirin

1,669

1.19 ± 0.10 (1)

410

1.37 ± 0.35 (1)

Salsalate

121

1.28 ± 0.34 (2)

107

1.30 ± 0.30 (2)

Ibuprofen

503

1.94 ± 0.43 (3)

238

2.34 ± 0.55 (3)

Naproxen

939

2.17 ± 0.23 (4)

327

3.43 ± 0.58 (4)

Sulindac

511

2.24 ± 0.39 (5)

220

2.89 ± 0.45 (5)

Piroxicam

790

2.52 ± 0.23 (6)

291

3.33 ± 0.46 (6)

Fenoprofen

161

2.95 ± 0.77 (7)

71

3.09 ± 0.65 (7)

Ketoprofen

190

3.45 ± 1.07 (8)

147

3.44 ± 0.78 (8)

Meclofenamate

157

3.86 ± 0.66 (9)

84

4.43 ± 0.84 (9)

Tolmetin

215

3.96 ± 0.74 (10)

120

4.83 ± 0.78 (10)

Indomethacin

386

3.99 ± 058 (11)

159

4.32 ± 0.60 (11)

These data comprise toxicities from all patients in the database and those that are considered “drug starts”. The results obtained with these differing periods of drug exposure were essentially similar. Ibuprofen was in a group with the two other salicylates, aspirin and salsalate) having the lowest toxicity ratings

An epidemiological safety investigation known as the Safety Profile of Antirheumatics in Long-term Administration (SPALA) was undertaken during the late 1980s to 1990 involving 30,000 rheumatic patients in participating centres in West Germany (N = 9), Switzerland (N = 3) and Austria (N = 4) (Brune et al. 1992). Of the 10 most-frequently prescribed NSAIDs (N = 36,147 prescriptions) ibuprofen was the second most-frequently prescribed drug after diclofenac, ranking it fourth in the overall total number of ADRs among the ten drugs. As shown in Table 14 ibuprofen was associated with the least number of reactions in the GI, liver and biliary, and “body as a whole systems”.
Table 14

Frequencies of adverse events (AE) in relation to numbers (no.) of four most-frequently prescribed NSAIDs in the SPALA study in rheumatic patients in West Germany, Austria and Switzerland

Frequencies of adverse events (no. AEs/no. prescriptions) and percentage individual AEs

Organ system classes

Diclofenac

Ibuprofen

Indomethacin

Acemetacin

No. of prescriptions

14,477

4,037

3,896

3,633

Gastrointestinal system (%)

14.10

11.20

15.90

19.10

Skin and appendages (%)

3.50

3.30

3.50

4.50

Central and peripheral NS (%)

2.50

3.00

7.90

4.80

Liver and Biliary system (%)

2.20

0.70

1.80

1.50

Body as a whole—general (%)

2.70

2.20

3.10

4.10

From Brune et al. (1992)

These two studies show that ibuprofen at prescription-level doses given to rheumatic patients is amongst the lowest toxicity ratings of frequently prescribed NSAIDs.

Outcomes from large-scale clinical trials

The studies with the coxibs conducted during the past decade were undertaken with large numbers of patients under modern standards of clinical investigation and with demanding requirements to establish safety in the GI, CV and other organ systems where serious adverse events with the NSAIDs often occur at low frequencies. Ibuprofen was used in a number of these studies as a “bench standard” in recognition of it being accepted as amongst the safest of all NSAIDs that is still widely used in rheumatologic and other musculo-skeletal conditions. These studies have afforded a useful and high quality source of data for assessment of the adverse reaction and general safety profile of ibuprofen from studies performed with rigorous quality controls and in a setting where the drug is being critically evaluated against competitors.

The individual adverse reactions in GI, CV and other organ systems are reviewed in detail in later sections. Here, the overall adverse reaction profiles for ibuprofen and comparator drugs are viewed in global sense employing outcome measures that are considered good indicators of overall patient and physician acceptability for safety and efficacy.

As previously mentioned the tolerability in adverse events in clinical trials conducted with the objective of assessing celecoxib in osteoarthritis and rheumatoid arthritis Moore et al. (2005) in clinical trials for comparing the occurrence of responses and adverse reactions from celecoxib with that of ibuprofen, diclofenac, naproxen, paracetamol and rofecoxib (Table 8). The data showed that adverse event discontinuation with ibuprofen following 12 or 24+ weeks of treatment were similar to those from diclofenac or celecoxib when either the number of events or the percentage of discontinuation is compared (Table 8). In the TARGET study, lumiracoxib 400 mg was compared with ibuprofen 2,400 mg and naproxen 1,000 mg taken for 52 weeks (Farkouh et al. 2004; Schnitzer et al. 2004). There were two major studies performed one addressing the CV events (Farkouh et al. 2004) and the data from ibuprofen sub-study is reviewed here in which this drug had lower or comparable safety with naproxen.

These studies from large scale clinical trails attest to the comparable rates for withdrawals from trials with ibuprofen and the coxibs. They show that the newer coxibs are neither more effective or less likely to produce adverse reactions leading to withdrawals from studies compared with that of ibuprofen.

Adverse events at non-prescription (OTC) dosages

A considerable number of studies have been reported comparing the adverse reactions from non-prescription (OTC) doses of ibuprofen with placebo, paracetamol (acetaminophen) or other analgesics. These studies have been performed using a variety of methodologies and study designs, some of which may have been critical to the outcomes.

Earlier reviews of published literature showed that OTC ibuprofen has comparable reports of adverse events (AEs) with paracetamol or placebo (Furey et al. 1993; DeArmond et al. 1995; Moore et al. 1996).

A systematic analytical review of published studies compared OTC ibuprofen with paracetamol where the drugs were taken as single doses or daily dosages up to 10 days (Rainsford et al. 1997; Tables 15, 16).
Table 15

Overall adverse event rates and exposure grouped by duration of dosing

Days dosed

Drug

No. of groups

Exposurea

Total number of patients

Overall percent with adverse events

Total number with adverse eventsb

Total number of adverse eventsc

<1

Paracetamol

27

0

4,644

10

444

479

<1

Ibuprofen

25

0

2,312

6

148

172

1

Paracetamol

11

420

420

10

43

49

1

Ibuprofen

5

215

215

8

18

22

2–7

Paracetamol

15

2,882

687

8

57

64

2–7

Ibuprofen

9

1,015

227

9

20

29

8–30

Paracetamol

6

5,496

207

19

39

39

8–30

Ibuprofen

9

5,960

272

19

52

52

31–90

Ibuprofen

5

6,504

85

29

25

29

Total

Paracetamol

59

8,798

5,958

10

583

631

Total

Ibuprofen

53

13,694

3,111

8

263

304

aNumber of patient days

bAdverse events grouped as the total number of patients having these events

cAdverse events grouped as the total of all recorded adverse events

From Rainsford et al. (1997)

Table 16

Statistical comparison of adverse events from data in Table 17 of studies where ibuprofen and paracetamol were compared directly (n ≥ 40, treatment ≥ 7 days)

Adverse events

F

P

Significance

Gastrointestinal

 Constipation

1.000

0.341

N/S

 Discomfort

1.225

0.297

N/S

 Vomiting

1.227

0.297

N/S

 Nausea

0.006

0.941

N/S

 Heartburn

0.001

0.975

N/S

 General

1.060

0.328

N/S

Gastrointestinal total

0.639

0.443

N/S

CNS total

1.013

0.338

N/S

Other total

1.138

0.311

N/S

Total

0.733

0.414

N/S

N/S, Not significant; analysis of variance single factor, level of significance set at 0.05

From Rainsford et al. (1997)

The subjects in these studies were either healthy volunteers, or those who had experienced various types of acute pain or chronic inflammatory conditions. Some studies involved comparisons with other analgesics/NSAIDs or placebo. Thus, there were a wide range of conditions in which the treatments were compared. The results showed that there were no significant differences between ibuprofen and paracetamol in occurrence of AEs after single or multiple daily doses taken for up to 10 days (Table 15) although there was a trend to increased GI AEs in both groups with increased duration of drug intake. There did not appear to be any discernable differences in AEs in different patient groups, although the number of patients in each of the groups was probably not sufficient to meet statistical requirements for being assessable. As paracetamol may be considered a ‘benchmark’ drug with a low propensity to cause serious GI events this study suggests as there was no differences in GI AEs between ibuprofen and paracetamol at OTC dosages, ibuprofen can be considered to have low risk of GI reactions comparable with paracetamol.

Kellstein et al. (1999) performed a meta-analysis of reports of randomized, double-blind, placebo-controlled parallel-group studies having initially reviewed published literature and established that only eight studies, all of which were unpublished but claimed as independent studies performed under the auspices of Whitehall–Robbins Healthcare met the criteria as specified above according to GCP conditions (Table 17). AEs were codified according to the conventional Coding Symbol Thesaurus for Adverse Reaction Terms (COSTART) with the exception of abdominal pain, which was “conservatively” assigned to the “body as a whole” digestive system. This may in fact have disguised the importance of this AE since it is a relatively frequent event in trials with NSAIDs and paracetamol.
Table 17

Number (N) and percentage (%) of subjects

Pool studies

All body systems

Digestive system

Body-as-a-whole system

Placebo N (%)

Ibuprofen N (%)

Placebo N (%)

Ibuprofen N (%)

Placebo N (%)

Ibuprofen N (%)

Single-day studies

7/318 (2.2)

1/319 (0.3)

2/318 (0.6)

1/319 (0.3)

5/318 (1.6)

0/319 (0.0)

Multiple-day studies

59/775 (7.6)

38/775 (4.9)

21/775 (2.7)

14/775 (1.8)

36/775 (4.6)

21/775 (2.7)

All studies

66/1,093 (6.0)

39/1,094 (3.6)

23/1,093 (2.1)

15/1,094 (1.4)

41/1,093 (3.8)

21/1,094 (1.9)

Experiencing a severe adverse reaction over all body systems, the digestive system, and the body-as-a-whole system

From Kellstein et al. (1999)

The eight selected studies were in mixed patient groups comprising three in OA pain, two in Delayed Onset Muscle Soreness (DOMS), and one each in sore throat pain, dental pain and a study of Maximal Use Safety and Tolerability (MUST) of non-prescription ibuprofen. The study dosages ranged from 400 mg b.i.d. (800 mg day−1, 2 studies) and 400 mg t.i.d. (1,200 mg day−1, 6 studies) with the duration of intake being 1–10 days. The primary purpose was to compare the effects of single-dose with multiple daily doses of ibuprofen with placebo. The subjects covered a wide range of ages (12–97 years) and racial groups of both genders in a total of 1,094 ibuprofen and 1,093 placebo-treated subjects.

Table 17 summarizes the serious AEs from this study. The principle outcomes can be summarized thus:
  1. (a)

    The overall number of AEs, those in body-as-a-whole and the digestive system were greater after multiple doses compared with single doses of ibuprofen and placebo.

     
  2. (b)

    There were no differences in AEs in all body systems and in the digestive system after single doses of ibuprofen compared with placebo or in the digestive system and body-as-a-whole after multiple doses.

     
  3. (c)

    In an analysis of individual AEs by COSTART, dizziness was identified among the central nervous system reactions to be significantly increased after multiple doses in the ibuprofen (2.5%) compared with placebo (1.4%); there being no differences after single doses of the treatments.

     
  4. (d)

    Overall tests for homogeneity among the study groups using the Breslow–Day statistical test showed no significant differences between the occurrences of all individual AEs over all the study groups.

     
  5. (e)

    Serious AEs over all categories were fewer in the ibuprofen compared with placebo in both the single and multiple dosage groups. Urinary tract infections while rare were more frequent in ibuprofen than placebo groups.

     

The reason for higher rates of AEs from placebo in “all body systems” and “body-as-a-whole” compared with ibuprofen is attributed to a larger number of patients in the placebo group reporting headaches, neck pain and malaise. The lower rates of these reactions in the ibuprofen groups are consistent with its analgesic activity.

While the studies employed in this meta-analysis are from unpublished investigations that have not been subjected to peer-review they are none-the-less from investigations that were performed according to GCP requirements and would have been in the company database that is subject to scrutiny by the US FDA.

Another study from the same company involved a prospective investigation of GI tolerability of the maximum daily OTC dose of 1,200 mg ibuprofen compared with placebo taken for 10 days in 1,246 healthy volunteers (Doyle et al. 1999). A total of 19% of ibuprofen-treated subjects (67 of 413) and 16% of placebo treated (161 of 833) individuals experienced GI AEs, there being no significant differences between the two groups. The GI adverse reactions were dyspepsia, abdominal pain, nausea, diarrhoea, flatulence and constipation. Occult blood tests were positive in 1.4% of all subjects; there being no differences between the two treatments in the occurrence of these reactions. The results in this prospective study confirmed the data from previous retrospective studies and showed that ibuprofen at OTC dosages has comparable GI reactions to placebo.

In a pharmaco-epidemiological investigation involving questionnaires distributed to 40 pharmacies in the Campania region of Southern Italy, Motola et al. (2001) observed that GI AEs were the most frequent with OTC and prescription NSAIDs with the incidence being 5.5% overall.

In a large scale general practice (known as the PAIN Study) based investigation in 4,291 patients in France, Le Parc et al. (2002) compared the tolerability of 7 days treatment of ibuprofen (up to 1.2 g day−1) with aspirin (up to 3 g day−1) for relief of musculoskeletal conditions. So-called “significant” AEs were reported in 15.0% of patients who took ibuprofen, 17% on paracetamol and 20.5% on aspirin; the difference between the ibuprofen and paracetamol groups being not statistically significant but significantly different from the aspirin group (Table 18). GI AEs were fewer in the ibuprofen group (4.4%) than in the paracetamol (6.5%) or aspirin (8.6%) group, the differences in all groups being statistically significant from one another. In the non-musculoskeletal group there were similar trends although there was no occurrence of serious digestive AEs.
Table 18

Most frequent significant adverse events by COSTART body systems and terms

Systems/terms

Ibuprofen (%)

Aspirin (%)

Paracetamol (%)

Body as a whole

5.4

7.4

5.7

Digestive system

3.6

4.7

4.3

Nervous system

1.1

2.2

1.1

Respiratory system

1.2

1.5

1.3

Abdominal pain

2.0

5.1

2.7

Nausea

1.6

1.8

1.3

Dyspepsia

0.9

1.9

1.3

Headache

1.2

1.1

1.6

From Moore et al. (2002)

Using the data acquired in the abovementioned PAIN Study, Moore et al. (2004) performed an assessment of risk factors that accounted for the development or association with AEs. By employing multivariant logistic regression analysis of 8,633 evaluable patients they identified the following risk factors (a) indication (e.g. musculo-skeletal pain, sore throat, colds and flu, menstrual pain, headache) (b) concomitant medications, (c) history of previous GI disorders, and (d) female gender. Age was not a risk factor. There were fewer clinically significant risk factors for GI AEs in the ibuprofen compared with paracetamol groups. The overall conclusion was that the main risk factor was concomitant medications.

A meta-analysis undertaken by Ashraf et al. (2001) in elderly (>65 years) osteoarthritic patients in which the incidence of adverse events (COSTART coded) from ibuprofen 1,200 mg daily was compared with those that received placebo. Following an initial assessment of the quality of papers, three independent clinical trials that had been performed by Whitehall Robins (USA) were selected in which the drug treatments were for ≤10 days. The pooled overall incidence of adverse events was 29.4% with the ibuprofen group (N = 197 patients) and 29.0% in the placebo group (N = 210 patients), with the three studies individually showing no statistically significant differences. The percentages of adverse events in the organ systems were for (a) ‘body as a whole’—12.7% on ibuprofen c.f. 9.5% on placebo; (b) digestive system—12.2% on ibuprofen c.f. 13.3% on placebo; and (c) nervous system—10.2% on ibuprofen c.f. 8.1% on placebo; the differences being not statistically significant.

This study is important in showing that ibuprofen at OTC doses is safe in the elderly OA patient, a group who frequently self-administer the drug.

Several investigations have been performed in what could be regarded as “at risk” patients (based on prior clinical history) either those admitted to hospitals for clinical investigation (and who could be regarded as being at “suspect” risk because of indicative symptoms requiring investigation) or patients with rheumatic diseases. The focus of these studies has been to identify the risks of serious AEs in the GI tract from intake of OTC analgesics. The rheumatic patients may have increased susceptibility to GI events from intake of NSAIDs as a consequence of their disease, concurrent disease (e.g. diabetes, CHD), concomitant medications (either anti-rheumatic e.g. steroids, or other agents to control diabetes, hypertension, CV disorders or subnormal renal function), as well as socio-psychologic stress or Helicobacter pylori infection. Since many patients with rheumatic disorders take OTC analgesic medications on a self-administered p.r.n. basis and as a reflection on costs of prescription NSAIDs which for the elderly or members of lower socio-economic classes could be a major problem, use of OTC analgesics in these patients can be regarded as one of the “real-world” uses of these drugs.

Among the reports in GI suspect risk patients Blot and McLaughlin (2000) reported investigations conducted by a mail survey of members of the American College of Gastroenterology (ACG) designed to identify risks of GI bleeding associated with intake of analgesics at OTC dosages within the previous week of intake. The methodology involved data collected from the ACG Registry (N = 627 patients) and “procedure-matched” endoscopy controls. Suspect factors (e.g. tobacco, alcohol intake etc.) were also identified. These hospitalized patients had a variety of upper or lower GI conditions that led to bleeding in the OTC analgesic group but with no bleeding in the control group. The number of patients in these groups might be considered relatively small and questions can be raised about the statistical validity of the sub group analysis of risk factors. The cases tended to be older subjects (mean 60 years) compared with controls (55 years) with 45% cases being over 65 years compared with 33% controls, and they were more often male (63%) cases compared with controls (49%). The balance of races was comparable with about two thirds being non-hispanic whites. GI risk, especially in the upper tract was greater in those that had consumed alcoholic beverages, this being increased in smokers, but cigarette smoking was unrelated to GI risks.

Of the major analgesics reported intake of drugs was associated with GI bleeding in 9.5% aspirin-takers, 4.2% ibuprofen takers, and 5.4% paracetamol users. A considerable number of patients had taken mixtures of two analgesics or prescribed NSAIDs. It should be emphasized that the numbers of patients were relatively small among the single analgesic-users (56 aspirin, 25 ibuprofen, 32 paracetamol) so it is questionable to ascribe causality to individual drugs. Non-the-less these data are instructive at least for assessment of potential GI bleeding in at risk patients. It is interesting that paracetamol was associated with GI bleeding as it is normally considered a low risk GI “safe” drug. In this complex group of patients with evident underlying disease it is clear that ibuprofen is somewhat safer than the other two analgesics.

In conclusion: (a) the studies at prescription-level doses show that ibuprofen has amongst the lowest risks for adverse events, (b) this drug is as good in safety and efficacy as any of the newer coxibs (which were designed to have lower incidence of adverse reactions), with serious events being rare, and (c) at OTC doses ibuprofen has low or at least amongst the lowest rating of risks for developing adverse reactions compared with other analgesics.

Adverse events and safety in paediatric populations

The safety profile of ibuprofen has been extensively evaluated in traditional paediatric clinical trials of fever and/or pain have been evaluated in a number of trials or in critical reviews (Walson et al. 1989; Czaykowski et al. 1994; Rainsford et al. 1997, Rainsford 1998, 1999, 2001; Diez-Domingo et al. 1998; van den Anker 2007). All have shown the low incidence of serious and non-serious adverse events (AEs) with ibuprofen. While these data are useful it is only in large scale population based studies that it is possible to obtain sufficient data to obtain a sound basis for safety evaluation.

In a series of papers, Lesko and Mitchell (1995; 1997; Lesko et al. 2002) have compared the safety of ibuprofen and paracetamol with a focus in particular on ibuprofen, in practitioner-based clinical trials, the methodologies of which were reviewed by Mitchell and Lesko (1995). These studies have been supported by both leading companies producing the antipyretics in the USA as well as the US FDA, US NIH and other pharmaceutical companies supporting the Sloane Epidemiology Unit of Boston University School of Medicine (Brookline, MA, USA) where these studies have been based. Among these studies the use of Advisory Groups has been employed which helps retain the abilities to critically assess data and ensure proper conduct of trials.

In their practitioner-based population study of 2,015 primary care physicians throughout continental United States of America, Lesko and Mitchell (1995) undertook a randomised, double-blind, office-based paracetamol (acetaminophen)-controlled trial of a total of 84,192 patients aged 6 months to 12 years of age who were randomly assigned to receive ibuprofen 5 or 10 mg kg−1 suspensions (Children’s Motrin®, McNeill), or 12 mg kg−1 paracetamol suspension (Calpol®, Burroughs Wellcome) for the treatment of acute febrile illness. The study provided for a 4 week follow-up period to determine the occurrence of side-effects. The primary outcome measures were hospitalizations for acute GI bleeding, acute renal failure and anaphylaxis. The occurrence of Reye’s syndrome was also monitored. Secondary outcomes included identification of previously unrecognized serious reactions. Two patients died one in a road accident who had received paracetamol and the other who had ibuprofen, died from bacterial meningitis; both these fatalities could be considered to be unrelated to the drugs. A total of 1% of the patients was admitted to hospital in the 4 weeks following enrolment. Four children were hospitalized for acute GI bleeding that was due to ibuprofen (2 from 10 mg kg−1 and 2 from 5 mg kg−1 of the drug) giving a risk of GI bleeding as 7.2 per 100,000 (95% CI 2–38 per 1000,000) with the risk from paracetamol being zero; the difference being not statistically significant. Gastritis or vomiting was observed in 20 patients that had received ibuprofen giving a risk of 36 per 100,000 (95% CI 22–55) and in 6 patients on paracetamol, the risk being 21 per 100,000 (95% CI 7.9–46). There were 24 patients who had received paracetamol who had asthma (RR = 85, 95% CI 55–150) and 44 on ibuprofen (RR = 80 per 100,000; 95% CI 57–110), thus showing there was no difference in risks between the two drugs. There was no risk from Reye’s syndrome, acute renal failure or anaphylaxis among 55,785 children that received ibuprofen. Low white blood cell count was observed in eight children that had received ibuprofen (but the causality could not be established) and none in the paracetamol group. The authors considered that the risks from outcomes of low incidence could not be ascertained because of the power of the study. This study attests to the low risks for serious GI, renal of anaphylactic events form ibuprofen, and a lack of association with severe renal or asthmatic events.

In what is probably the largest study designed to investigate the safety of analgesics in children ≤2years old, Lesko and Mitchell (1999) used data from the Boston Collaborative Fever Study in a total of 27,065 febrile children who were randomized to receive 5 or 10 mg kg−1 ibuprofen of 12 mg kg−1 paracetamol suspensions. The study was double-blind and practitioner-based with children being eligible if, in the opinion of the attending physician, their illnesses warranted treatment with an antipyretic; duration and height of fever were not a criteria for participation. Follow-up was achieved by mailed questionnaire or telephone interviews. The most common cause of fever in children was otitis media (45%), upper respiratory tract infection (40%), pharyngitis (15%), lower respiratory tract infection (7.4%) and gastro-intestinal infection (2.2%). Data from the two doses of ibuprofen were combined because there were no discernible differences between the groups; thus the size of the ibuprofen group is about twice that of the paracetamol group.

The risk of hospitalization for any reason in the 4-weeks after enrolment (N = 385 patients in total) was the same in the ibuprofen group [Relative Risk, RR = 1.1 (0.9–1.3) compared with that of the paracetamol group as a reference RR = 1.0]. The absolute risks were 1.5% (1.3–1.6, 95% CI) for the ibuprofen group and 1.4% (1.1–1.6%) for the paracetamol group. None of the study participants was hospitalized for acute renal failure, anaphylaxis or Reye’s syndrome. Three children who received ibuprofen were hospitalized for GI bleeding; which was non-serious and was resolved with conservative management. The risk of hospitalization from GI bleeding was estimated to be 11 per 100,000 (95% CI 2.2–32 per 100,000) for antipyretic assignment and 17 per 100,000 (95% CI, 3.5–49 per 100,000) in those children ≤2 years who received ibuprofen. Among children <6 months of age there was no observed risk for hospitalization for any of the primary outcomes.

The risks with hospitalization for asthma/bronchiolitis for ibuprofen were 0.9 (95% CI, 0.5–1.4) compared with paracetamol; a total of 65 children being hospitalized for these group of conditions (they were grouped together because of frequent misdiagnosis of these two conditions). Of nine children hospitalized for vomiting or gastritis the risk did not vary according to antipyretic assignment.

Of the 385 who were hospitalized, those in whom creatinine levels were available (29%) and were considered to be only of borderline statistical significance between the ibuprofen and paracetamol groups. There was no significant differences between these two treatment groups when age, weight, sex or admission diagnosis of dehydration were compared. When alternate cut-off points were used to define an elevated creatinine level (44 or 53 μmol L−1) there were no significant differences between the antipyretic groups.

While this was the largest controlled study ever undertaken of antipyretic use in children ≤6 months of age the authors admitted that the power to detect serious adverse events is limited (especially those that occur at low frequency). Some clinical and demographic information suggested that the study participants probably reflected a wide spread of febrile illnesses in the view of the authors even though socioeconomic data were limited.

These data are important in showing that (a) there is a remarkable low incidence of serious and even non-serious ADR’s in children ≤2 years and especially those ≤6 months which receive antipyretic therapy for febrile illness.

Another large investigation into the overall safety of ibuprofen in paediatric populations was performed by Ashraf et al. (1999). This study known as the Children’s Analgesic Medicines Project (CAMP) was a prospective, multicentre, all-comers, multi-dose, open-randomized and open label study designed to compare the safety of ibuprofen (Children’s Advil®) with that of paracetamol (Children’s Tylenol®) given for relief of pain and/or fever. A total of 41,810 children aged 1–18 years were enrolled in a so-called naturalistic outpatient paediatric setting (PEGASUS Research Inc., Salt Lake City, UT, USA) involving 68 clinics in the USA. Among 30,144 children who took one dose of either ibuprofen or paracetamol, 14,281 were “younger” aged ≤2 years and 15,863 were “older” and aged 2–12 years. There were no serious AEs in ≥1% of patients in either group. There were no cases of Reye’s syndrome, gastric bleeding and/or ulceration, renal failure, necrotizing fasciitis, Steven’s Johnson or Lyell’s syndromes, anaphylaxis or any other serious condition that are known to be associated with either drug in any population.

Statistically significant but small and clinically in-significant differences were observed in AEs in both age groups that received ibuprofen compared with those that had paracetamol being 17.6% c.f. 15%, respectively in the younger and 11.9 and 10.7% in the older groups. The increased incidence of AEs in the ibuprofen groups was related to the greater disease severity in those groups. Four deaths were recorded (herpes encephalitis, sepsis due to Staphyloccocus pneumoniae, medulloblastoma and sudden death syndrome) and were unrelated to the study medications. Those AEs related to the special senses occurred occasionally followed by the digestive and respiratory systems and skin (all in 3–4% approximately in the younger and slightly lower in the older group).

Specific organ toxicity

Gastro-intestinal (GI) toxicity

Serious GI ADRs (upper GI bleeding and ulcers) are a major cause of concern and in the past 3–4 decades have aroused much interest among clinicians, experimentalists and regulators alike (Voutilainen et al. 1998; Wolfe et al. 1999; Lewis et al., 2005; Schaffer et al. 2006; Arroyo and Lanas 2006; Lanas et al. 2006; Laine et al. 2006). The problems are particularly apparent in rheumatic patients (Singh et al. 1996; Singh 2000) and the elderly (Griffin 1998; Beyth and Schorr 1999; Seinelä and Ahvenainen 2000; Mamdani et al. 2002; Kean et al. 2008). Definitions vary on what constitutes the elderly but most agree >65, a time that seems to have derived over a century ago from Otto von Bismark who required Prussian officers to retire at this age) (Kean and Buchanan 1987; Buchanan 1990). Early studies indicated that ibuprofen was well-tolerated in elderly patients (Buckler et al. 1975).

A range of factors influence the development of NSAID-associated GI ulcerations and bleeding (Table 19). This makes it difficult to ascribe a quantitative component of the NSAID to the occurrence of serious GI events.
Table 19

Risk factors for the development of NSAID-associated gastro-duodenal ulcers

Established risk factor

Possible risk factor

Advancing age

High Alcohol consumption

High dose NSAID or paracetamol

Cigarette smoking

Use of more than 2 NSAIDs

Helicobacter pylori Infection

Concurrent paracetamol

 

Concurrent anti-coagulants

 

Concurrent aspirin (even low dose)

 

Prior history of peptic ulcer disease

 

Rheumatoid arthritis

 

From Wolfe et al. (1999) and Laine (2001); modified and with additional information from Rainsford (1984; 2004a, 2005b)

A number of studies have reported that prescription dosage level ibuprofen produces time and dose-dependent blood loss (assessed using the radiochromium blood loss technique) from the GI tract of volunteers or patients (Warrington et al. 1982; Aabakken et al. 1989; Hunt et al. 2000) and mild-moderate endoscopic changes in fasted human volunteers (Lanza et al. 1979, 1981, 1987; Friedman et al. 1990; Bergmann et al. 1992; Roth et al. 1993; Müller et al. 1995; Gallego-Sandin et al. 2004) or those with rheumatic diseases (Teixeira et al. 1997). The extent of the loss of blood maybe overestimated using the radiochromium technique as a consequence of biliary excretion of the radiolabelled chromium (Schneider et al. 1984; Rainsford 2004a). The extent of the mucosal changes (lesions, ulcers) and blood loss from ibuprofen is low compared with other NSAIDs but is above that of placebo and paracetamol (Strom et al. 1997; Rainsford 1999c).

Epidemiological studies of GI risks

A considerable number of population studies have been reported over the past 2–3 decades comparing the occurrence of serious GI events from ibuprofen and other NSAIDs, at prescription dosage level dosages, with some studies being dose-ranging (Kaufman et al. 1993; Langman et al. 1994; Henry et al. 1993, 1996; MacDonald et al. 1997; Henry and McGettigan 2003; Hippisley-Cox et al. 2005; Thomsen et al. 2006; see also Table 20). The study designs, outcomes measures and variables (dosage and duration) vary considerably among these studies. Some measures having included the occurrence of peptic ulcer bleeds (PUBs), upper GI bleeding, ulcers viewed at endoscopy (usually investigated as a consequence of clinical symptoms or as part of a planned investigation) or the more general grouping of “serious events”. While these studies vary considerably they are useful in comparing the risks of serious GI events attributed to ibuprofen with that of a range of other NSAIDs with known ulcerogenicity.

A summary of some of the population studies reported in the 1990s in the period before the introduction of the newer class of coxib NSAIDs is shown in Table 20.

In a meta-analysis of published studies comparing the GI ADRs for various NSAIDs, Henry et al. (1996, 1998) was able to show that the relative risks of these events from different NSAIDs ranged considerably (Fig. 7). They found that ibuprofen had the lowest risks for developing GI complications (Fig. 7).
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig7_HTML.gif
Fig. 7

From Henry et al. (1998). Reproduced with permission of Kluwer Academic Publishers (now part of Springer Verlag)

https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig8_HTML.gif
Fig. 8

From Henry et al. (1998). Reproduced with permission of Kluwer Academic Publishers (now part of Springer Verlag)

Henry et al. (1996, 1998) also observed (a) dose-related occurrence of GI complications with ibuprofen, naproxen and indomethacin (Fig. 7), and (b) the ranking of GI complications was directly related to the plasma elimination half-life (t1/2) of the individual NSAIDs (Tables 20, 21 and Fig. 8). Again, as with the overall analysis ibuprofen had the lowest rates of occurrence of GI complications which is attributed to its short t1/2 (~2 h). Thus, there is good pharmacokinetic rationale to account for the low GI ADRs with ibuprofen.
Table 20

Serious outcome gastro-intestinal toxicity ranking of tNSAIDs

Drug

Kaufman et al. (1993)

Henry et al. (1993)

Langman et al. (1994)

Rodrigues and Jick (1994)

Henry et al. (1996)

MacDonald et al. (1997)

Aspirin

    

10

 

Azapropazone

  

1

 

1

2

Diclofenac

6

7

6

7

9

4

Diflunisal

 

1

 

8

7

 

Fenbufen

     

11

Fenprofen

4

  

5

11

1

Ibuprofen

7

8

7

9

12

8

Indomethacin

5

3

4

4

5

9

Ketoprofen

1

2

2

2

2

5

Nabumetone

     

10

Naproxen

3

4

5

3

6

6

Mefenamic acid

     

7

Piroxicam

2

5

3

1

3

3

Sulindac

 

6

 

6

8

 

Tolmetin

    

4

 

Toxicity rankings of NSAIDs with those associated with the greatest risk of ulcer complication is given the number 1. The studies used different methodologies

Table 21

Ranking of GI complications from NSAIDs with plasma elimination Half-Life (t1/2) of the drugs

G-I safety, dose and plasma half-life of NSAIDS

 Ranking of RR of ulcers c.f. t1/2 (h)

  Ibuprofen (2.5) < Diclofenac (1.5–5) < Diflunisal (10.8) < Fenprofen (2.2) < Aspirin (0.5–4.5) < Sulindac (14.0) < Naproxen (14.0) < Indomethacin (3.8) < Piroxicam (48.0) < Ketoprofen (8.5) < Tometin(6.8) < Azapropazone (22.0)

 Dose-relationships—low c.f. high dose:

  Ibuprofen RR 1.8–4.0

  Naproxen RR 3.8–6.0

  Indomethacin RR 2.3–6.5

From Henry et al. (1998)

Among the most comprehensive studies that have been performed to evaluate overall adverse drug reactions in European populations has been the study by Lugardon et al. (2004). These authors undertook an analysis of spontaneous reports to the French Pharmacovigilance network which is probably one of the most extensive and comprehensive pharmacovigilance systems in Europe. A summary of the data shown in the Table 22 compares the reporting odds ratios for GI events of heteroarylacetic acids that include ibuprofen, diclofenac, naproxen and ketoprofen with that of the two principle coxibs rofecoxib and celecoxib as well as the oxicams, principally meloxicam, piroxicam and tenoxicam. The unadjusted and adjusted odds ratios for ibuprofen ADRs are amongst the lowest all the drugs that were studied by Lugardon et al. (2004).
Table 22

Adverse drug GI reaction reporting odds ratio (OR) (with their 95% confidence interval) according to main classes of non-steroidal anti-inflammatory drugs (NSAIDs)

Drugs

Adjusted ORa (95% CI)

Adjusted ORb (95% CI)

Coxibs

4.6 (3.3–6.5)*

14.9 (9.3–23.7)*

Rofecoxib

5.2 (3.1–8.7)*

21.0 (10.6–41.6)*

Celecoxib

3.7 (2.4–5.8)*

11.7 (6.6–20.9)*

Oxicams

12.2 (6.7–22.2)*

25.3 (11.9–53.6)*

Heteroaryl acetic acids

Ibuprofen

4.5 (2.3–8.8)*

7.3 (3.2–16.6)*

Diclofenac

3.9 (2.1–7.2)*

9.2 (3.8–22.2)*

Naproxen

10.6 (4.7–23.7)*

17.9 (6.7–47.6)*

Ketoprofen

8.6 (5.3–13.9)*

19.9 (10.7–37.0)*

aAdjustment for matching factors (age, sex, period of occurrence)

bAdjustment for matching factors (age, sex, period of occurrence) and confounding factors (regional pharmacovigilance centre, work place of health professional and drug-exposure (anticoagulants, anti-platelet drugs, aspirin, gastroprotective agents, and other NSAIDs)

From Lugardon et al. (2004)

* P < 0.0001

A similar conclusion can be drawn from the case–control study of Laporte et al. (2004) as shown in Table 23.
Table 23

Gastro-intestinal bleeding from NSAIDs in a multicentre case–control study in Spain and Italy

Drug

Cases (no (%))

Controls (no (%))

Odds ratio (95% CI)

Population attributable risk (%)

NSAIDs

 Aceclofenac

15 (0.5)

30 (0.4)

1.4 (0.6, 3.3)

 Aspirin (acetylsalicylic acid)

591 (21.1)

403 (5.7)

8.0 (6.7, 9.6)

18.5

 Dexketoprofen

16 (0.6)

8 (0.1)

4.9 (1.7, 13.9)

0.5

 Diclofenac

100 (3.6)

98 (1.4)

3.7 (2.6, 5.4)

2.6

 Ibuprofen

60 (2.1)

58 (0.8)

3.1 (2.0, 4.9)

1.5

 Indomethacin

29 (1.0)

16 (0.2)

10.0 (4.4, 22.6)

0.9

 Ketoprofen

16 (0.6)

9 (0.1)

10.0 (3.9, 25.8)

0.5

 Ketorolac

33 (1.2)

6 (0.1)

24.7 (8.0, 77.0)

1.1

 Meloxicam

14 (0.5)

11 (0.2)

5.7 (2.2, 15.0)

0.4

 Naproxen

52 (1.9)

27 (0.4)

10.0 (5.7, 17.6)

1.7

 Nimesulide

48 (1.7)

46 (0.6)

3.2 (1.9, 5.6)

1.2

 Piroxicam

119 (4.3)

40 (0.6)

15.5 (10.0, 24.2)

4

 Rofecoxib

10 (0.4)

10 (0.1)

7.2 (2.3, 23.0)

0.3

 Other NSAIDs

34 (1.2)

33 (0.5)

3.6 (2.0, 6.8)

0.9

 NSAIDs + antiplatelet drugs

140 (5.0)

54 (0.8)

16.6 (11.3, 24.2)

4.7

Analgesics

 Lysine Clonixinate

26 (0.9)

47 (0.7)

1.3 (0.7, 2.6)

 Metamizole

117 (4.2)

155 (2.2)

1.9 (1.4, 2.6)

2

 Paracetamol (acetaminophen)

376 (13.4)

612 (8.6)

1.2 (1.0, 1.5)

 Propyphenazone

17 (0.6)

38 (0.5)

1.3 (0.6, 2.8)

From Laporte et al. (2004)

Case–control investigations by Garcia-Rodriguez and Hernandez-Diaz (2001) using data from the UK General Practice Database also show the low risks of GI events with ibuprofen in contrast with those from various doses and periods of taking NSAIDs with or without concurrent aspirin (Table 24). These data are also instructive in highlighting the fact that high dose paracetamol (hitherto regarded as a GI safe drug) when taken at doses of >2 g day−1 alone or in combination with NSAIDs is associated with relative risks >2 (alone) or >6 (combination with NSAIDs) of haemorrhage.
Table 24

Epidemiological data from general practice database (UK) on peptic ulcer bleeding risks from aspirin and other NSAIDs and paracetamol and other sources

Drug

Usage/factor

 

Relative risk

Ibuprofen

 

Lowest risk (dose-dependent)

~1.0–2.0

Aspirin

Overall use

Users

2.0

cf Non-users

1.0

Recent users

1.5

Past users

1.1

Dose

75–300 mg day−1

2.1

>400 mg day−1

3.1

<50 mg day−1

0.7

Period of use

1–60 day

4.5

61–180 day

2.7

181–730 day

1.0

>730 day

1.6

Paracetamol

 

<1 g day−1

1.0

 

1–2 g day−1

0.9

 

2–4 g day−1

3.4

 

>4 g day−1

6.5

 

2 g with NSAID

4.2

 

>2 g with NSAID

13.5

 

cf NSAID alone

3.5

NSAIDs

 

Low dose

2.5

 

High dose

5.0

Duration

1–30 days

4.3

 

>730 days

3.5

Formulation

Plasma t1/2

 
 

<12 h (high dose)

4.2

 

≥12 h

5.4

 

(slow release

6.2

 

<12 h (low dose)

2.4

 

≥12 h

2.8

Data from Garcia-Rodriguez and Hernandez-Diaz (2001)

In rheumatic patients, Singh (2000) has produced data from the ARAMIS (a US rheumatic disease patient database) showing relatively high risks of GI bleeding (or peptic ulcer bleeds) from all NSAIDs, with little difference between individual NSAIDs (including ibuprofen). Lower risks were associated with paracetamol. While there have been claims made by this author that the GI risks are similar to those from OTC dosages of these analgesics there is little information available on their duration of use, concomitant medications and other risk factors.

More insight into the GI risks associated with OTC analgesics/NSAIDs has been provided by Lewis et al. (2005) in a case–control study of hospitalized patients recruited from 28 hospitals. The cases (N = 359) had upper GI bleeding, benign gastric outlet obstructions or perforations, while controls (N = 1,889) were obtained from random-digit phone dialling in the same region. Use of OTC doses of non-aspirin NSAIDs ≥4 days in the past week was associated with an adjusted odds ratio of 1.83 (95% CI 1.14–2.95); the risks from ibuprofen being much lower. Risks were increased with higher doses of the drugs confirming what has been well-known about the dose–response relationships among most NSAIDs being associated with serious GI AEs (Henry et al. 1996, 1998).

GI effects of ibuprofen in coxib studies at prescription doses

Data on the GI peptic ulcer bleeds (PUB), bleeding ulcers or endoscopically observed upper GI injury have been obtained from a considerable number of large scale clinical trials comparing the effects of coxibs with ibuprofen and other NSAIDs at prescription doses. These data are useful for giving and indication under controlled clinical trial conditions in arthritic populations of the relative risks of developing serious gastro-duodenal injury at high doses of these drugs.

Among these studies the data on ulcer complications in the CLASS study observed at 6 months showed there were differences between celecoxib and NSAIDs (Silverstein et al. 2000). However, as pointed out by Jüni et al. (2002) these differences were not apparent at 12 months (Table 25) suggesting there are time-dependent factors that are significant in considering ulcer incidence of both coxibs and NSAIDs. The clinical significance of these data like that from other long term studies is that when coxibs are taken for relatively short periods of time (2–4 weeks) they are less likely to cause ulcer complications than NSAIDs such as naproxen and diclofenac especially if they are taken for several months or longer. There is also the issue of what has been described as “channelling” where patients with a history of GI complaints or GI ulcer disease may be prescribed coxibs in the belief they will be “gastric-safe”, this may be such that benefits for using these dugs may be less apparent and the patients may require anti-ulcer co-therapy (e.g. with H2 receptor antagonists or proton pump inhibitors [PPIs]). The cost-benefits of coxibs therapy may prove less favourable as not only are these drugs notably more expensive than conventional NSAIDs but if PPIs or other anti-ulcer therapies have to be employed they may as well be given with cheaper NSAIDs, especially those with a lower propensity to cause CV complications (e.g. naproxen) or combinations with aspirin for cardioprotection.
Table 25

Summary of adverse events in the CLASS study

Event %

Celecoxib

Diclofenac

Ibuprofen

GI

45.6

55.0

46.2

 Withdrawal

12.2

16.6*

13.4

Renal

6.8

6.7

10.3*

 Withdrawal

1.0

0.6

1.3

CV non-ASA

1.6

1.2

0.4

Hepatic

1.8

6.9*

1.9

 Withdrawal

0.3

3.5*

0.3

Based on data published by Silverstein et al. (2000)

* Significantly different compared with celecoxib

These calculations (Table 26) show that although the percentage of relative risk reduction (RRR) for celecoxib c.f. NSAIDs (ibuprofen, diclofenac) is 61% and for rofecoxib c.f. naproxen is 60% the values for absolute risk reduction (ARR) are relatively small being 0.7 and 0.8%, respectively (Shoenfeld 2001). The latter values represent the fact that the percentage incidence of GI complications for the NSAID as well as for the two coxibs is in the low end range of 0.4–1.14% which are very low percentages. Thus, with such small differences calculations of RRR are meaningless and give a false impression of improved benefit to the GI tract of the coxibs. This approach of using RRR percentage benefits has been extensively exploited in published data on coxib trials and must, therefore, be regarded as suspect statistical treatment of data which has little relevance clinically. Indeed clinical significance in many coxib trials has rarely been considered in contrast to statistical significance.
Table 26

Estimation of serious gastrointestinal reactions from the CLASS trial of celecoxib compared with NSAIDs

In the CLASS Trial (non-aspirin-using patients only)

 Percentage of patients with serious NSAID-associated gastrointestinal complications was:

 Celecoxib = 0.44% diclofenac = 0.48% (no statistically significant difference between Diclofenac and celecoxib)

Celecoxib = 0.44% Ibuprofen = 1.14%

 ARR = 1.14 − 0.44% = 0.7%

 NNT = 1/0.7% = 1/0.007 = 143

 RRR = 1.14% − 0.44%/1.14% = 61%

Results are reported as serious NSAID-associated gastrointestinal complications (i.e. gastrointestinal bleeds, perforations and obstructions) per 100 patient-years

ARR = absolute risk reduction; RRR = relative risk reduction

NNT = number needed to treat

From Schoenfeld (2001)

In another large scale coxib study Sikes et al. (2002) compared the incidence of gastric and duodenal ulcers from two dose levels of valdecoxib 10 and 20 mg day−1 with that from ibuprofen 2,400 mg day−1 and diclofenac 150 mg day−1 taken for 12 weeks. This study was of shorter duration than the other coxib studies reviewed previously. There is a trend noted by Jüni et al. (2002) for differences in GI events between celecoxib and the NSAIDs to become smaller with time. Thus at the shorter time period it might have been expected that if valdecoxib had a favourable GI profile that it would show lower incidence of GI events compared with the two comparator NSAIDs. The data in Table 27 shows that there was a somewhat higher incidence of gastric and duodenal ulcers and all GI ulcers from ibuprofen compared with valdecoxib, but these were fewer than seen with diclofenac.
Table 27

Incidence rate of upper-gastrointestinal ulcers from ibuprofen compared with diclofenac and valdecoxib

 

Placebo

Valdecoxib 10 mg daily

Valdecoxib 20 mg daily

Ibuprofen 800 mg t.i.d

Diclofenac 75 mg b.i.d.

12-week cohort (n (%))

 N

123

142

157

149

145

 Gastroduodenal

8 (7)

7 (5)

7 (4)

24 (16)a,b,c

25 (17)a,b,c

 Gastric

6 (5)

5 (4)

6 (4)

20 (14)a,b,c

20 (14)a,b,c

 Duodenal

3 (2)

2 (1)

1 (1)

6 (4)

7 (5)c

ITT cohort (n (%))

 N

178

189

198

184

187

 Gastroduodenal

8 (4)

7 (4)

7 (4)

25 (14)a,b,c

25 (13)a,b,c

 Gastric

6 (3)

5 (3)

6 (3)

21 (11)a,b,c

20 (11)a,b,c

 Duodenal

3 (2)

2 (1)

1 (1)

6 (3)

7 (5)c

12-week cohort includes patients who took the study medication for the entire 12-week period. ITT cohort includes all patients who had a post-treatment endoscopy irrespective of whether they completed 12 weeks of treatment

From Sikes et al. (2002)

aSignificantly different from placebo at P > 0.05

bSignificantly different from valdecoxib 10 mg daily at P > 0.05

cSignificantly different from valdecoxib 20 mg daily at P > 0.05

This study is of importance in that there were separate analysis of ulcer incidence in (a) H. pylori negative and H. pylori positive subjects, and (b) between those who were aspirin takers (for CV prophylaxis) and non-takers.

The data in Table 28 shows that H. pylori status made little if any difference to the ulcer incidence in subjects that received any of the drugs, but taking of aspirin did increase the incidence of ulcers in the ibuprofen and naproxen groups and to a lesser extent in the valdecoxib groups.
Table 28

Week 12 gastroduodenal ulcer incidence (%) by aspirin use, age or Helicobacter pylori status

Group

Placebo

Valdecoxib 10 mg daily

Valdecoxib 20 mg daily

Ibuprofen 800 mg t.i.d

Diclofenac 75 mg b.i.d.

H. pylori-positive

4/46 (9)a,b

3/40 (8)a,b

2/49 (4)a,b

9/45 (20)

11/54 (20)

H. pylori-negative

4/119 (3)a,b

3/134 (2)a,b

5/136 (4)a,b

15/123 (12)

14/122 (12)

H. pylori status unknown

0/2 (0)

1/3 (33)

0/6 (0)

0/7 (0)

0/5 (0)

Taking aspirin

0/26 (0)a,b

3/16 (19)a,b,c

2/29 (7)a,b

10/31 (32)c

10/33 (30)c

Not taking aspirin

8/141 (6)a,b

4/161 (3)a,b

5/162 (3)a,b

14/144 (10)

15/148 (10)

Age ≥ 65 years

4/65 (6)a,b

6/56 (11)a,b,d

3/67 (5)a,b

15/71 (21)d

14/78 (18)

Age > 65 years

4/102 (4)a,b

1/121 (1)a,b

4/124 (3)a,b

9/104 (9)

11/103 (11)

From Sikes et al. (2002)

aP ≤ 0.014 vs. ibuprofen

bP ≤ 0.008 vs. diclofenac

cP ≤ 0.017 vs. not taking aspirin

dP ≤ 0.025 vs. age < 65 years

From the point of view of GI safety, there may have been pathological consequences of hepato-renal ADRs and hypertension that contributed to the vascular aetiology of upper GI ulcer disease as well as the consequences of the diuretics and anti hypertensive drugs (which as noted earlier increase the risk for developing ulcers) as well as NSAID-drug interactions that patients with hepato-renal conditions and hypertension received for treatment of these conditions (Rainsford et al. 2008). In the end, what has emerged in the safety analysis of the coxibs is summarized in Table 26 it is clear that the benefits of what now are classed as “first generation” coxibs (celecoxib, rofecoxib, valdecoxib) may have been marginal compared with some conventional NSAIDs, among which etodolac, ibuprofen, nabumetone and possibly diclofenac (although the intestinal ulceration and hepatotoxicity reduce any favourable a favourable safety profile it might have had).

An interesting and possibly important point that should be considered is the arthritic condition in the CLASS as well as other studies with the coxib. Patients in the VIGOR only had RA half of whom were receiving corticosteroids but were not allowed aspirin at cardioprotective doses (Silverstein et al. 2000) whereas those in the CLASS study had both RA and OA and were allowed aspirin (Bombardier et al. 2000). It has been claimed that there were no differences in ulcer complications in patients with OA c.f. RA which is surprising in view of the differing medications allowed in these studies. This is in one sense surprising since as noted earlier it has been speculated that patients with RA may be more susceptible to NSAIDs than those with OA. It could be that the selection criteria for patients entered in the CLASS Study were such that RA as well as OA patients were relatively “fit” and without complicating chronic conditions that inevitably occur in older more infirmed patients, especially those with RA (Kean et al. 2008).

Indeed the incidence of ulcers in the CLASS study (Tables 25, 26) as well as in the meta analysis of celecoxib trials by Moore et al. (2005) reveal a remarkably low incidence in placebo and NSAID groups. This gives support to the view that patients selected for inclusion in these studies may have been of relatively better health. Another issue is that if there were any real differences in data concerning ulcer complications in say a proportion of patients with RA c.f. those with OA these may have been disguised in the grouping of data together such as in the CLASS results.

In conclusion, the epidemiological and large-scale clinical trials show that ibuprofen has amongst the lowest risk of NSAIDs for serious GI events. The concomitant ingestion of aspirin may raise the risk of GI complications in a similar way to that seen with celecoxib and rofecoxib (Strand 2007).

GI symptomatic adverse reactions

GI symptoms (nausea, heartburn, epigastric distress, vomiting, diarrhoea) are among the major reasons for withdrawal by patients from therapy with NSAIDs. Much effort has been undertaken to establish if there is a relationship between symptoms and gastric injury but in general the results have been inconclusive. Yet often symptoms are a cause for referral for endoscopic or other clinical investigations.

Meta-analysis of the tolerability and adverse events from a range of trials of celecoxib compared with NS-NSAIDs, paracetamol and placebo (using data from the published and unpublished trials from Pfizer) (Moore et al. 2005) revealed some interesting features and trends concerning the occurrence of GI symptoms notably nausea, dyspepsia, diarrhoea, abdominal pain, vomiting. These constitute main reasons (other than ulcer/bleeds or other serious ADRs) for withdrawal from therapy, and indeed the data by Moore et al. confirmed this pattern.

Data on the GI symptoms from NSAIDs and coxibs summarised in the report by Moore et al. (2005) highlight that (1) the occurrence and relative risks of most GI symptoms in patients receiving celecoxib, rofecoxib or paracetamol being greater than that of placebo, (2) while there are trends for a lower incidence of some symptoms with low dose celecoxib the differences are less distinct with higher dose celecoxib, and (3) the data on confidence intervals in relative risks with most comparisons often overlaps to the extent that it is doubtful if any differences, especially those favouring celecoxib, have any meaning.

The authors of this study noted that the proportion of patients having dyspepsia was about 7% and that there were no differences in comparison with placebo, paracetamol or rofecoxib, but there were more patients on NS-NSAIDs. Celecoxib was responsible for abdominal pain in about 5% patients there being no difference c.f. placebo or paracetamol but more patients on NS-NSAIDs and rofecoxib experienced this adverse effect. Other GI symptomatic effects as well as overall GI tolerability there were trends in favour of celecoxib in comparison with the other treatments but the 95% confidence intervals for relative risk often overlapped those of comparator drugs. This makes extrapolation of these findings difficult from what are relatively small values for incidence and percentage differences of symptomatic GI ADRs like that of clinical ulcers and bleeds in this meta analysis (Moore et al. 2005) is probably of limited value.

GI events at OTC dosages

GI symptoms (nausea, epigastric or abdominal pain, dyspepsia, diarrhoea, flatulence and constipation) are among more frequent reactions observed with OTC use of ibuprofen as well as with paracetamol and aspirin, and generally the symptoms are of the same order as in subjects who have received placebo (Rainsford et al. 1997, 2001; Doyle et al. 1999; Kellstein et al. 1999; Ashraf et al., 2001; Le Parc et al. 2002; Boureau et al. 2004; Biskupiak et al. 2006). The occurrence of GI symptoms with ibuprofen has often been found to be lower than with aspirin and comparable with those from paracetamol (Rainsford et al. 1997; Moore et al. 1999; Le Parc et al. 2002; Boureau et al. 2004). Serious GI reactions are rare and have not been reported in significant numbers in trials with OTC ibuprofen (Doyle et al. 1999; Kellstein et al. 1999; Ashraf et al. 2001; Le Parc et al. 2002; Boureau et al. 2004). Thus, it may be concluded that GI events are essentially non-serious with OTC ibuprofen, are probably reversible upon cessation of the drug (an action likely to be taken by most subjects), and are no different from those with paracetamol and less than with aspirin.

GI safety in paediatric populations

As noted earlier in the discussion of the large-scale paediatric study by Lesko and Mitchell (1995) four children were hospitalized for acute GI bleeding that was due to ibuprofen (2 from 10 mg kg−1 and 2 from 5 mg kg−1 of the drug) which gives a risk of GI bleeding as 7.2 per 100,000 (95% CI 2–38 per 1000,000) with the risk from paracetamol being zero; the difference was not statistically significant. Gastritis/vomiting was observed in 20 patients that had received ibuprofen with a risk of 36 per 100,000 (95% CI 22–55) and in 6 patients on paracetamol with a risk of 21 per 100,000 (95% CI 7.9–46).

In their later study in ≤2 year olds, Lesko and Mitchell (1995) observed that three children who received ibuprofen were hospitalized for GI bleeding. The risk of hospitalization from GI bleeding was estimated to be 11 per 100,000 (95% CI 2.2–32 per 100,000) for antipyretic assignment and 17 per 100,000 (95% CI, 3.5–49 per 100,000) in those children ≤2 years who received ibuprofen.

As noted earlier in the discussion of the large-scale paediatric study by Ashraf et al. (1999) no occurrences of gastric bleeding or ulcers were observed with either ibuprofen or paracetamol. The incidence of adverse events (AEs) in the digestive system was 3.0 and 2.1% in the younger group (≤2 years) that received ibuprofen or paracetamol and 2.1 and 1.2% for these drugs in the older group (2–12 years); the statistical tests showed the former being non-significant but in the latter this was statistically significant. Abdominal pain occurred in 0.6% of the younger patients that had ibuprofen compared with 0.1% that had paracetamol, while in the older group the incidence was 0.6 and 0.2% for ibuprofen and paracetamol, respectively.

These results attest to the relative gastric safety in children of ibuprofen in comparison with paracetamol accord with earlier investigations (Walson et al. 1989; Czaykowski et al. 1994; Rainsford et al. 1997; Diez-Domingo et al. 1998). They confirm the safety of both drugs being comparable and relatively low in the open clinical paediatric study.

Cardiovascular safety

There are three main issues concerning cardio-vascular (CV) safety of ibuprofen. The first of these concerns the possible risks of triggering serious CV conditions such as congestive heart failure (CHF) and myocardial infarction (MI); a situation which has arisen as a consequence of the re-evaluation of risks of MI from all NSAIDs following the identification of risks of this condition with rofecoxib and other coxibs (Purcell 2007; Strand 2007; Layton et al. 2008; Solomon et al., 2008; Sørensen et al. 2008; van der Linden et al. 2008). The second issue concerns the effects of NSAIDs, including ibuprofen, on blood pressure in hypertensive individuals; elevation of blood pressure being regarded as a potential marker for risks of MI or stroke (Topol 2004, 2005; White 2007). This situation is complicated because most NSAIDs reduce the effectiveness of anti-hypertensive drugs and diuretics as a consequence of blunting by NSAIDs of these drugs on their prostaglandin-mediated actions (Hersh et al. 2007; Ishiguro et al. 2008; White 2008). Linked to the effects of NSAIDs in elevating blood pressure are their effects on renal functions which can contribute to their hypertensive potential as a consequence of inhibition of renal prostaglandin production (White 2007). This has given rise to the so-called “cardio-renal” syndrome of NSAIDs and again has come from recognition of the pronounced renal effects of coxibs as a consequence of inhibition of COX-2 in the macula densa (Harris et al. 1994; Haas et al. 1998; Khan et al. 1998; Inoue et al. 1998; Ichihara et al. 1999; Wolf et al. 1999; Roig et al. 2002); aspects of this are discussed under “Renal Toxicity”. The third issue is the possibility that ibuprofen might reduce the anti-platelet effects of aspirin and thus reduce the anti-thrombotic effectiveness of the latter (Purcell 2007).

Serious CV conditions

The available evidence reviewed here shows that ibuprofen has low CV risks although there may be effects on blood pressure and on the actions of drugs used to control blood pressure.

Serious CV events principally ischaemic heart conditions were initially highlighted by long term studies with the coxibs (Topol 2004, 2005; Kanna et al. 2005; Ostor and Hazleman 2005; Rainsford 2005a; White 2007). The CV events included myocardial infarction and hypertension and were noted particularly with rofecoxib (Vioxx®) in the VIGOR study as well as in a number of other studies. They were of sufficient concern for the company producing this drug, Merck Sharp & Dohme, to withdraw it from the market on 30th September 2004 (Topol 2004, 2005; American College of Rheumatology, Hotline, 2005; Psaty and Furberg 2005). In the wake of the issues surrounding withdrawal of rofecoxib the FDA determined that valdecoxib (Bextra®, Pfizer Inc.) had similar CV risks and this as well as the skin reactions that emerged with this drug lead to its withdrawal in 2005.

The FDA and other agencies worldwide were alerted and alarmed about the CV ADRs with rofecoxib such that extensive reviews were undertaken by these agencies world-wide of both coxibs and NSAIDs based on the somewhat unfounded premise that inhibition of COX-2 which occurs with all these drugs might well underlie the increased risks of MI and elevation of blood pressure. It is important to note that the VIGOR study investigating the long-term GI effects of rofecoxib was performed in patients with rheumatoid arthritis (RA). RA patients are known to have a markedly higher risk of developing MI and other serious CV events (Nurmohamed et al. 2002; Assous et al. 2007) and this is not a feature generally recognized in the assessment of CV risks of coxibs and NSAIDs. Furthermore, there are indications from the studies performed with the coxibs in long-term preventative studies in cancer or Alzheimer’s disease that high doses of these drugs were employed and these patients were clearly very sick.

Etoricoxib (Arcoxia®) also developed by Merck Sharp and Dohme is probably the most selective inhibitor of COX-2 of those drugs that have been developed to date. While long-term investigations are awaited there are indications that GI and CV events from etoricoxib may be lower than with NSAIDs and celexocib.

Lumiracoxib (Prexige®; Novartis) is not chemically like that of other coxibs as it is a derivative of diclofenac. It also does not have the high COX-2 selectivity of etoricoxib or rofecoxib. Novartis embarked on large scale studies to determine the CV safety with lumiracoxib which was compared with ibuprofen and naproxen all at prescription level dosages (Schnitzer et al. 2004; Matchaba et al. 2005; Farkouh et al. 2009) (see Tables 30, 31).

Epidemiological studies

The awareness of CV risks from coxibs and NSAIDs has led to a substantial number of studies have been reported in which the risks of MI or other serious CV accidents have been examined. Among these Garcia-Rodriguez et al. (2004) employed data in the UK General Practice Research Database (GPRD) which records reports from GP’s sent anonymously to the UK MHRA. This database records demographic and patient data and over 90% of GP referrals along with prescription details. The associations of MI with various patient and risk factors were discriminated in this study. The odds ratios (ORs) for development of MI after multi-variant adjustment with current use of NSAIDs were found to have an OR = 1.06 (0.87–1.29; 95% CI values) for ibuprofen, contrasted with those at the upper end of risk with piroxicam with an OR of 1.25 (0.69–2.25). Prior history of CHD or concomitant intake of aspirin did not increase the risk of MI from ibuprofen.

Jick et al. (2006) performed a case–control analysis of data from the UK GPRD did not find any increased risk of acute MI with either ibuprofen or naproxen, but did find increased risks with diclofenac, rofecoxib and celecoxib.

Kimmel et al. (2004) performed a study of hospitalized patients for MI in a five-county region around Philadelphia (USA). They observed reductions in the risks of MI among non-aspirin NSAID users. The reduction was also observed with ibuprofen with the adjusted OR being 0.52 (0.39–0.69; 95% CI) compared with that of 0.53 (0.42–0.67) for aspirin and 0.48 (0.28–0.82) for naproxen, a drug which has often been found to have low CV risk and in fact may have some cardio-protective actions (Topol 2004, 2005; Kanna et al. 2005; Kean et al. 2008).

In a nested case–control study of data from a leading US health maintenance organization, Kaiser Permanente, Graham et al. (2005) examined 8,143 cases of serious coronary disease (from 2,302,029 patient-years follow-up). They found that the adjusted OR for ibuprofen for current use was 1.26 (1.00–1.60), compared with that of naproxen OR = 1.36 (1.06–1.75) and rofecoxib low dose (< 25 mg day−1) OR = 1.47 (0.99–2.17) and high dose (>25 mg g−1) OR = 3.58 (1.27–10.11). For remote use the OR for ibuprofen was 1.06 (0.96–1.17) contrasted with that of rofecoxib (high dose, >25 mg day−1) with an OR of 3.0 (1.09–8.31).

In a combined study of CV and GI events in 49, 711 US Medicare beneficiaries (>65 years of age) Schneeweiss et al. (2006) found that the risks of acute MI was 1.20 with ibuprofen, compared with 1.01 with naproxen, 1.54 with diclofenac, 1.58 with celecoxib and 1.56 with rofecoxib.

Another nested case–control study using clinical records of the UK general practice database (known as QRESERCH) Hippisley-Cox and Coupland (2005) found that recent use (<3 months) of ibuprofen was associated with an adjusted OR of 1.24 (1.11–1.39), compared with that of diclofenac which had an OR of 1.55 (1.39–1.72), naproxen 1.09 (0.96–1.24), celecoxib 1.14 (0.93–1.40) and rofecoxib 1.05 (0.89–1.24). The lack of any signals with rofecoxib and to some extent with celecoxib is odd but may reflect more limited use of this drug according to the guidelines by the UK National Institute for Clinical Excellence (NICE).

A retrospective study of hospitalization records of ≥65-year-old patients admitted for acute MI (as well as GI events) in Québéc (Canada) by Rahme and Nedjar (2007) showed that the adjusted hazard ratios for ibuprofen were 1.05 (0.74–2.41), 1.69 (1.35–2.10) for diclofenac, 1.59 (1.31–1.93) for naproxen, 1.34 (1.19–1.52) for celecoxib, 1.27 (1.13–1.42) for rofecoxib, and 1.29 (1.17–1.42) for paracetamol.

Other recent studies have essentially confirmed the view that tNSAIDs (including ibuprofen) have risks for CV events similar to those of some of the coxibs but the data on risks are quite variable (Strand 2007; Layton et al. 2008; Solomon et al. 2008; Sørensen et al. 2008; van der Linden et al. 2009).

Overall, therefore, these epidemiological investigations highlight (with admittedly some variability of ORs) the relatively low-moderate risk of ibuprofen being associated with serious CV conditions such as MI. These observations contrast with the higher risks with diclofenac, the coxibs, and in one study with paracetamol, and variable risks with naproxen.

Clinical trials and meta-analyses

The CLASS study (Table 25) showed that ibuprofen had a relatively low incidence of cardiovascular events and this has been confirmed in a number of other studies with coxibs. These large scale studies are valuable for highlighting under control conditions in clinical trials in patients with rheumatic disease that ibuprofen has a low risk of developing cardiovascular effects.

The most recent evaluation of the cardiovascular risks with coxibs and NSAIDs undertaken by Antman et al. (2007) which was a study under the auspices of the American Heart Association has been instructive in critically evaluating data and offering a clear statement of cardiovascular risks of all NSAIDs including coxibs. As shown in Table 29 (which is a summary of the cardiovascular risk reported in placebo-controlled clinical trials with the non-selective NSAIDs) the various outcome measures and nature of assessments from either randomized controlled trials observational studies or registry data showed that there is some variability in cardiovascular risk with the different NSAIDs. Overall, ibuprofen has a slightly lower relative risk than diclofenac but naproxen has a notably lower risk of cardiovascular events. Indeed, in the VIGOR study and several other studies that were reviewed back in 2004 it was found that naproxen had the lowest overall risk of all NSAIDs for developing CV events. For ibuprofen the relative risks range from 1.07 for all CV, mostly MI, to all vascular events 1.51. It should be noted, however, that the confidence intervals for these risks overlap considerably and are approximately unity in comparison with placebo.
Table 29

Cardiovascular risks of ibuprofen and other NSAIDs from meta-analysis of placebo-controlled clinical trials, observation studies and registry data

Type of study

Outcome

RR

95% CI

Versus placebo or no treatment

 Naproxen

   

  Meta-analysis of RCTs

Vascular events

0.02

0.67–1.26

  Meta-analysis of OSs

CV events, mostly MI

0.07

0.87–1.07

 Ibuprofen

   

  Meta-analysis of RCTs

Vascular events

1.51

0.06–2.37

  Meta-analysis of OSs

CV events, mostly MI

1.07

0.07–1.16

  Registry

Recurrent MI

1.25

1.07–1.46

  Registry

 

1.50

1.36–1.67

 Diclofenac

   

  Meta-analysis of RCTs

Vascular events

1.63

1.12–2.37

  Meta-analysis of OSs

CV events, mostly MI

1.40

1.16–1.70

  Registry

Recurrent MI

1.54

1.23–1.43

  Registry

 

2.40

2.00–2.80

Versus selective COX-2 inhibitor

   

 Naproxen

   

  Meta-analysis of any non-naproxen NSAID (primarily diclofenac or ibuprofen)

Vascular events

0.64

0.49–0.83

  Meta-analysis of RCTs indicates randomized, controlled trails

Vascular events

1.14

0.80–1.46

From Antman et al. (2007)

OSs observational studies, CV cardiovascular, MI myocardial infarction

A review in the form of a memorandum from Dr. David J. Graham of the Office of Drug Safety of the FDA (addressed to Dr. Paul Siegelman, acting Director of Office of Drug Safety) entitled “Risk of Acute Myocardial Infarction and Sudden Cardiac Death in Patients treated with COX-2 selective and non-selective NSAIDs”, dated 30th September 2004 (see http://www.fda.gov/.../Drugs/DrugSafety/PostmarketDrugSafetyInformationforPatientsandProviders/ucm1068; accessed 16 Nov 2009), showed that ibuprofen had a risk of acute MI of 1.09 (CI 95% 0.9–1.21) compared with that of naproxen relative risk 1.18 (CI 95% 1.04–1.35) whereas that of rofecoxib ranged from 3.15 to 1.29 in a dose-related manner.

In the large multicentre trial intended to establish the risks of CV events from lumiracoxib with those from ibuprofen or naproxen in some 14,000 patients it was found that the number of confirmed or probable myocardial infarctions and ischaemic events in the ibuprofen group was comparable with that from lumiracoxib as well as the naproxen treatment group (Fig. 9; Tables 30, 31; Farkouh et al. 2004, 2009).
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig9_HTML.gif
Fig. 9

Incidence of confirmed or probable myocardial infarctions (clinical and silent), from ibuprofen compared with lumiracoxib and naproxen by sub-study and aspirin use. From Farkouh et al. 2004)

Table 30

Incidence of confirmed or probable ischaemic events including myocardial infarctions (clinical and silent), from ibuprofen compared with lumiracoxib and naproxen by sub-study and aspirin use

 

Both sub-studies

Lumiracoxib versus ibuprofen sub-study

Lumiracoxib versus naproxen sub-study

Lumira-coxib

NSAIDs

Lumira-coxib

Ibuprofen

Lumira-coxib

Naproxen

Number of patients in non-aspirin population

6,950

6,968

3,401

3,431

3,549

3,537

 Patients with confirmed or probable ischaemic events (%)

34 (0.49)

27 (0.39)

13 (0.38)

12 (0.35)

21 (0.59)

15 (0.42)

 All myocardial infections (%)

14 (0.20)

9 (0.13)

4 (0.12)

5 (0.15)

10 (0.28)

4 (0.11)

  Clinical (%)

14 (0.20)

5 (0.07)

4 (0.12)

3 (0.09)

10 (0.28)

2 (0.06)

  Silent (%)

0

4 (0.06)

0

2 (0.06)

0

2 (0.06)

 Ischaemic stroke (%)

12 (0.17)

8 (0.11)

6 (0.18)

2 (0.06)

6 (0.17)

6 (0.17)

 Unstable angina (%)

5 (0.07)

5 (0.07)

1 (0.03)

4 (0.12)

4 (0.11)

1 (0.03)

 Transient ischaemic attack (%)

3 (0.04)

5 (0.07)

2 (0.06)

1 (0.03)

1 (0.03)

4 (0.11)

Number of patients in aspirin population

2,167

2,159

975

966

1,192

1,193

 Patients with confirmed or probable (%)

29 (1.34)

24 (1.11)

9 (0.92)

9 (0.93)

20 (1.68)

15 (1.25)

 Ischaemic events

 All myocardial infarctions (%)

9 (0.42)

8 (0.37)

1 (0.10)

2 (0.21)

8 (0.67)

6 (0.50)

  Clinical (%)

6 (0.28)

7 (0.32)

1 (0.10)

2 (0.21)

5 (0.42)

5 (0.42)

  Silent (%)

3 (0.14)

1 (0.05)

0

0

3 (0.25)

1 (0.08)

 Ischaemic stroke (%)

11 (0.51)

9 (0.42)

2 (0.21)

4 (0.41)

9 (0.76)

5 (0.42)

 Unstable angina (%)

5 (0.23)

6 (0.28)

3 (0.31)

3 (0.31)

2 (0.17)

3 (0.25)

 Transient ischaemic attack (%)

4 (0.18)

1 (0.05)

3 (0.31)

0

1 (0.08)

1 (0.08)

From Farkouh et al. (2004)

Table 31

Combined incidence of gastrointestinal and cardiovascular events from lumiracoxib, compared with ibuprofen and naproxen, by sub-study (safety population (Schnitzer et al. 2004, 2009)

Both substudiesa

Number of patients with events/number at risk (%)

Hazard ratio (95% CI)

P*

Lumiracoxib

89/9,117 (98%)

0.65 (0.49–0.84)

0.0014

Non-steroidal anti-inflammatory drugs

133/9,127 (1.46%)

  

Lumiracoxib vs ibuprofen substudyb

 Lumiracoxib

30/4,376 (0.69%)

0.50 (0.32–0.79)

0.0025

 Ibuprofen

56/4,397 (1.27%)

  

 Lumiracoxib versus naproxen substudyb

   

 Lumiracoxib

59/4,741 (1.24%)

0.75 (0.53–1.05)

0.0961

 Naproxen

77/4,730 (1.63%)

  

Cox proportional-hazards models include, in addition to treatment group, the factors

aSub study, low-dose aspirin, and age

bLow-dose aspirin and age

* Based on Wald χ2 statistic for treatment group comparison

The combination of the risks of CV and GI events has been considered a major element in determining the overall safety of coxibs and NSAIDs (Antman et al. 2007). To examine this Schnitzer et al. (2004) undertook an analysis of the combined risks of these drugs. A summary of their data showing the combined GI and CV events is shown in Table 31 (Schnitzer et al. 2004). These data show that the hazard ratio is significantly lower for lumiracoxib than the two NS-NSAIDs although there is considerable overlap of the values of the 95% confidence intervals. These data do show, however, that the combined risks of serious CV and GI events with ibuprofen are relatively low.
Table 32

Patterns of hepatopathy observed with various hepatotoxic agents

Drug

Exposed cases

Total

Cholestatic

Hepatocellular

Mixed

Anti-bacterial agents

 Clavulanic acid

17

9

3

5

 Isoziazid + rifampicin + pyrazinamide

8

1

7

 Isoniazid

4

3

1

 Erythromycin

3

1

2

Analgesics

 Paracetamol

17

4

11

2

Musculo-skeletal system

 Acetylsalicylic acid

10

2

7

1

 Diclofenac

5

1

4

 Allopurinol

3

2

1

 Droxicam

3

1

1

1

 Nimesulide

3

3

 Aceclofenac

2

1

1

 Indomethacin

2

2

Alimentary tract and metabolism

 Ranitidine

8

4

3

1

 Glibenclamide

5

3

2

 Omeprazole

5

1

2

2

 Ebrotididne

4

4

 Metoclopramide

4

1

3

Cardiovascular system

 Ticlopidine

7

7

 Atenolol

5

1

2

2

 Amiodarone

2

1

1

From Sabate et al. (2007)

Overall, therefore, these studies show that ibuprofen has low risks for developing cardiovascular events, principally serious condition such as myocardial infarction although vascular events might be slightly increased in risk.

Interaction of ibuprofen with the anti-platelet effects of aspirin

The possibility that ibuprofen may interfere with the anti-platelet effects of aspirin was highlighted by the article by Catella-Lawson et al. (2001) in the New England Journal of Medicine. They undertook a study which aspirin (81 mg) was taken 2 h before ibuprofen (400 mg) each morning for 6 days. The order of taking the two drugs was reversed and a similar design was incorporated with paracetamol 1,000 mg. Serum thromboxane B2 levels as an indicator of COX-1 activity in platelets and platelet aggregation were measured and were found to be significantly inhibited by aspirin; the maximum inhibition being evident on day 6 when the drug was taken alone. The authors found that when aspirin was given followed by ibuprofen as well as before taking aspirin then there was complete inhibition of the effect of aspirin on serum thromboxane and platelet aggregation. This impairment of platelet aggregation and thromboxane production by ibuprofen was not evident with paracetamol, diclofenac or rofecoxib.

The consequence of these studies was that there were a considerable number of pharmaco-epidemiological investigations to establish if NSAIDs would in general impair the anti-thrombotic potential of aspirin and its prevention of myocardial infarction. Thus, MacDonald and Wei (2003) analysed data from the Scottish Administrative Pharmacy Database and found that patients with cardiac disease had been prescribed combinations of ibuprofen and aspirin had an increase in cardiovascular mortality compared with that of patients who had taken aspirin alone. This effect was not evident when diclofenac was taken with aspirin.

About the same time Kurth et al. (2003) published data from patients enrolled in the Physicians Health Study which was a randomized trial in which subjects received aspirin and NSAIDs and who were at increased risk of cardiovascular events compared with patients who did not use NSAIDs. The increased risk of NSAIDs causing possibility of adverse events when given with aspirin while dose-dependent was relatively small and required the drugs to be used for long periods. The study by Kimmel et al. (2004) which has already been mentioned was interesting because this managed to put a completely different slant on the whole story. In patients with no history of coronary artery disease the use of aspirin was associated with the lower risk of myocardial infarction as expected but this benefit was not seen in patients who took any NSAIDs in addition to aspirin. Patients, who had established coronary disease, who used aspirin with NSAIDs were at similar risk of developing myocardial infarction compared with that of patients who had taken aspirin alone. Thus, there is an important issue relating to whether patients have coronary disease or not in the effect of the NSAIDs. It should be noted that the earlier study of Catella-Lawson et al. (2001) had been undertaken in normal subjects.

In a study in elderly patients, who had already experienced a myocardial infarction the mortality of those who had received aspirin and a non-steroidal drug was similar to that of patients who been prescribed alone (Ko et al. 2002; Curtis et al. 2003; Kean et al. 2008). No apparent differences were observed in the mortality and analysis of patients who had been prescribed aspirin and ibuprofen compared with those prescribed aspirin alone (Curtis et al. 2003).

A consensus view suggests that the mode of action of ibuprofen in impairing the inhibition by aspirin of platelet function is due to competition between ibuprofen and the active site of COX-1 which is irreversibly inhibited by covalent modification by the acetyl group of aspirin at or near the active site (Curtis and Krumholz 2004; Gaziano and Gibson 2006).

A pharmacodynamic study by Cryer et al. (2005) investigated the effects of ibuprofen on aspirin-induced thromboxane B2 production, which was intended as a follow-up to the paper by Catella-Lawson et al. (2001). This study was undertaken in 51 volunteers in a double-blind randomized parallel placebo-controlled study. The objections to the Catella-Lawson study were that it did not feature a placebo group and there were issues about the study population. The basis of the use of single measurement of thromboxane production is that this correlates to a high degree to the inhibition of platelet aggregation when aspirin is taken. Thromboxane production was measured over 10 days at 1, 3 and 7 days (in the period prior to randomization to treatment with ibuprofen or placebo) during 8 days treatment with 81 mg aspirin once daily in the morning. This resulted in greater than 90% thromboxane inhibition. On the ninth day and subsequently for 10 days the subjects were randomly assigned to receive ibuprofen or placebo and their thromboxane B2 levels were measured on day 0, 1, 3, 7 and 10. In both groups, there was greater than or equal to 98% inhibition of thromboxane B2 production although there was a small but clinically in-significant difference between the two treatment groups of thromboxane inhibition on day 7. Since this was already in a group who had greater that 98% mean inhibition of thromboxane production this could not be regarded as clinically significant (Cryer et al. 2005). These results show that prior treatment for 8 days with aspirin is not affected by subsequent ibuprofen treatment in terms of platelet thromboxane production. A similar study has been performed by Pongbhaesaj et al. (2003) and published in abstract form which showed almost identical results.

However, more recent investigations involving ex vivo production of COX-1 activity, measurement of platelet aggregation and modelling of data have indicated that ibuprofen at a single dose of 400 mg can cause a transient interference with the anti-platelet effects of aspirin but that with complete recovery after 6 h (Hong et al. 2008).

Thus, it may be concluded that the timing of aspirin and ibuprofen intake may have considerable bearing on the interaction of ibuprofen with aspirin on platelets. The clinical significance of this in terms of the prevention of cardiovascular disease in patients especially those taking OTC ibuprofen, who are at risk of developing these conditions clearly is of minor importance when viewed in context of the study by Kimmel et al. (2004).

Another important aspect arising from these studies is that ibuprofen itself inhibits platelet aggregation or functions (Brooks et al. 1973; McIntyre et al., 1978; Barclay 2005). The mechanisms of the inhibition of platelet aggregation by ibuprofen are, however, different from those of aspirin. Thus, Brooks et al. (1973) observed that 4 weeks treatment of male volunteers with ibuprofen 1,800 mg day−1 reduced aggregation induced by collagen and ADP but not in re-calcificated prior-citrated blood (a thrombin-induced reaction that is inhibited by aspirin). Forty minutes after 7 days treatment with ibuprofen platelet aggregation was inhibited but this returned to normal after 24 h; a situation where it would normally be expected that aspirin would have produced >90% inhibition of aggregation and prolongation of bleeding time. Moreover, ibuprofen does not cause inhibition of coagulation in re-calcificated prior-citrated blood or prothrombin times (Brooks et al. 1973).

Notwithstanding the obviously differing basis of the aspirin–ibuprofen interaction the US FDA pronounced a warning on the concomitant use of aspirin and ibuprofen in patients, who may be taking aspirin for the prevention of coronary vascular disease (Ellisson et al. 2007). Indeed, the FDA has published on its MedWatch website (http://www.fda.gov/medwatch/report.htm2007) information for health care professionals and drug facts concerning ibuprofen warning of the concomitant use of ibuprofen and aspirin. In the information for healthcare professionals it is stated that with occasional use of aspirin there is likely to be a minimal risk from any attenuation of the anti-platelet effects of low-dose aspirin because of the long-lasting effect of aspirin on platelets. Moreover, they state that patients who use immediate release aspirin (not enteric-coated) and take a single dose of ibuprofen 400 mg should take the dose of ibuprofen at least 30 min or longer after the aspirin to avoid attenuation of the effect of aspirin on platelets. They state that recommendations about the timing of concomitant use of ibuprofen and enteric-coated low-dose aspirin cannot be made on the base on available data. Thus, on the basis of information of the FDA and the available published literature it is clear that separation of the dose of aspirin from that of ibuprofen is a practical means of being able to avoid the potential for impairment of the anti-platelet effect of aspirin by ibuprofen.

It should be noted that an earlier study in patients with rheumatoid arthritis by Grennan et al. (1979) showed that high-dose aspirin (3.6 g day−1) but not a lower dose of 2.4 g day−1 in combination with high- or low-dose ibuprofen showed that there was a weak clinical additive effect on indices of articular function and pain and this appeared to be related to an increase in serum ibuprofen by aspirin but ibuprofen administration did not affect serum salicylate levels. Thus, high doses of aspirin (not those usually used for anti-thrombotic effects) may have some impact on the clinical efficacy of ibuprofen in a positive sense but this is related to effects on ibuprofen concentration in the plasma.

Effects of ibuprofen in hypertension

Elevation of blood pressure is regarded as an indicator or surrogate for CV risk especially in patients that are at risk of CV events. Studies with the coxibs, especially rofecoxib, indicated that they could increase blood pressure and produce oedema in patients with rheumatic conditions (Topol 2004, 2005; Kanna et al. 2005; Ostor and Hazleman 2005; Rainsford 2005a, b; Antman et al. 2007) NSAIDs, including ibuprofen, cause little or no increase in blood pressure in normotensive individuals (Pope et al. 1993; Johnson et al. 1994; Miwa and Jones 1999; Nurmohamed et al. (2002). This has been confirmed in extensive meta-analyses of various clinical trials (Johnson et al. 1994). The issue is, however, that NSAIDs interfere with the actions of β-blockers (Johnson et al. 1994) and other drugs (Fendrick et al. 2008). These effects are due to the influence of NSAIDs, including iburprofen, on production of prostaglandins and nitric oxide which can affect the actions of anti-hypertensive agents (Murray and Brater, 1999; Rodriguez et al. 2001; Roig et al. 2002). In a controlled clinical trial in patients with mild to moderate hypertension receiving anti-hypertensive medications (β-blockers + diuretics) 3 weeks treatment with ibuprofen 1,200 mg day−1 caused an increase in supine blood pressure by 5.3 mmHg and in sitting mean arterial pressure of 5.8 mm Hg compared with placebo (Radack et al. 1987). Similarly, increased blood pressure was noted in a placebo-controlled clinical trial in patients receiving hydrochlorthiazide and 1,800 mg day−1 ibuprofen (Gurwitz et al. 1996). However, in a study in stage 1 and 2 hypertensive patients on low and high sodium diets receiving the angiotensin-converting enzyme (ACE) inhibitor, enalapril, ibuprofen 1,200 mg day−1 did not affect systolic or diastolic blood pressure although in a related study indomethacin reduced the effects of capropril (Velo et al. 1987). Other NSAIDs are well-known to interfere with the actions of ACE inhibitors (Badin et al. 1997). Conversely, inhibition of the renin-angiotensin system upregulates COX-2 (Wolf et al. 1999) and thus may exacerbate NSAID related renal functions. Calcium channel blockers do not appear to be affected by ibuprofen and other NSAIDs in hypertensive patients (Miwa and Jones 1999).

Congestive heart failure and cardio-renal effects

Several studies have indicated that use of NSAIDs in patients with a history of heart disease may cause an increased risk of congestive heart failure (McGettigan and Henry 2000). It appears that this effect of NSAID intake maybe a class effect and confined to patients that have been taking the normal anti-arthritic doses of these drugs. The risk of increased occurrence of coronary heart failure is overall has an odds ratio of 2.8 but those with a history of heart disease this may be increased to 10.5. It appears that plasma half-life of elimination plays a role in the risk of coronary heart failure inasmuch as this risk seems to be doubled in long half-life versus short half-life drugs (McGettigan and Henry 2000).

Since inevitably the interference by NSAIDs by prostaglandin-dependent processes including haemostasis, vasodilatation, vasoconstrictor balance and renal functions (including electrolyte balance) influences the potential for cardiac toxicity via renal effects this class effect with NSAIDs usually seen at high doses of NSAIDs with long half-lives maybe a significant feature in the increase in hypertension and subsequent risk of cardiovascular disease (McGettigan et al. 2000).

In paediatric studies there have been no records of serious or non-serious CV or cardio-renal effects of ibuprofen (Ashraf et al. 1999).

Renal toxicity

Renal effects of ibuprofen are common to all those syndromes that are known to be produced by NSAIDs (Dunn et al. 1984; Brater 1998; Breyer 1999; Murray and Brater 1999; Mounier et al. 2006). The four main primary types of renal impairments observed with the NSAIDs include (a) acute ischaemic renal insufficiency, (b) effects on sodium potassium and water homeostasis with interference with the effects of diuretics and anti-hypertensive therapy and, (c) acute interstitial nephritis and renal capillary necrosis. The association of ibuprofen intake with the development of adverse renal effects is probably due to its widespread use rather than to any particular characteristic of the drug per se since irreversible effects are rare (Murray and Brater 1999). Renal dysfunction may be more pronounced in patients that have known risk factors including prior renal disease or impaired renal function (for example changes in creatinine clearance) (Chen et al. 1994; Bennett 1997; Castellani et al. 1997; Galzin et al. 1997; Schwartz et al. 2002). The issue is of probably greater concern in elderly subjects because of the higher prevalence of arthritic disease among them and the greater need for NSAID therapy (Murray and Brater 1999; Kean et al. 2008). Rarely serious renal pathology has been observed with ibuprofen (Carmichael and Shankel 1985; Radford et al. 1996; Cook et al. 1997; Silvarajan and Wasse 1997; Brater 1998; Murray and Brater 1999) but these reactions have not been observed in trials with OTC ibuprofen (Whelton et al. 1990; Whelton 1995). The evidence from literature surveys and clinical trials suggests OTC use of ibuprofen does not cause significant renal injury (Rainsford et al. 1997, 2001; Doyle et al. 1999; Kellstein et al. 1999; Hersh et al. 2000a; Ashraf et al. 2001; Le Parc et al. 2002; Boureau et al. 2004).

Griffin et al. (2000) using Tennessee (USA) Medicaid US Federal State Program Database with patients at greater than 65 years of age, undertook an analysis of the effects of NSAIDs on the development of acute renal failure in these elderly patients (see also Kean et al. 2008). Their analysis included consideration for conventional population variables as well as the concomitant intake of prescription drugs and aspirin. In their study they identified 1,799 persons aged greater that 65 years of age with a “community-acquired” pre-renal failure or intrinsic renal failure that required hospitalization for varying periods of time. Patients with acute renal failure are different from controls in a variety of population-related and drug-related factors. NSAIDs featured in the increasing risk for pre-renal failure among those without any underlying renal insufficiency. The odds-ratios were estimated (along with the 95% confidence intervals) for the risk of association between use of individual NSAIDs and hospitalization for acute renal failure among this population in a case–control comparison. Ibuprofen had an odds ratio of 1.63 (1.23–2.08 95% CI) and was exceeded only by piroxicam, fenoprofen and several other single or multiple-use NSAIDs. The slightly higher risk associated with ibuprofen intake may be a reflection of its wide-spread unsupervised use.

In relation to OTC use of ibuprofen and its possible association with the development of nephrotoxicity adverse reactions it is been noted that analgesic nephropathy is not been widely recognized or reported effect of renal OTC ibuprofen (Mann et al. 1993) and certainly this is only infrequently reported in ADR reports made to the UK CSM (Prescott and Martin 1992). In the analysis of risks of renal side effects of ibuprofen by Mann et al. (1993), it was found that the renal effects are dose-dependent and these effects are almost exclusively in elderly subjects with low intravascular volume and low cardiac output. Furey et al. (1993) observed that renal-vascular effects of OTC ibuprofen in elderly patients with mild thiazide-treated hypertension and renal insufficiency does not appear to be a risk factor for the development of renal compromise or hypertension even though renal function declines with age.

Farquhar et al. 1999; Farquhar and Kenney 1999, have shown that OTC dose of ibuprofen (1.2 g day−1) in normal subjects subjected to the heat-stress, low-sodium diet or dehydration may cause impairment of renal blood flow, glomerular filtration and electrolyte excretion which is related to inhibition of renal prostaglandin production. Studies in rabbits suggest that those with pre-existing renal failure receiving ibuprofen may have alterations in the pharmacokinetics of the two enantiomers of the drug (Chen et al. 1994) with the clearance of the active (S+) isomer being significantly impaired in this model of renal dysfunction. Thus, there may be increased prostaglandin inhibition in the renal tubular systems in individuals with renal impairment.

Overall, these studies suggest that OTC ibuprofen is a low risk factor for developing acute or chronic renal conditions but that as with other NSAIDs there is increasing risk particularly in elderly individuals or those with compromised renal function where the drug is taken at high prescription anti-arthritic doses (Kean et al. 2008).

Renal effects in children

A number of nephrologists, paediatricians and other physicians have expressed concerns over the years about the risks of serious analgesic-related renal reactions in children (van den Anker 2007).

In the two large-scale practitioner-based studies as part of the Boston University Fever Study, Lesko and Mitchell (1995, 1999) noted in particular that there were no cases of acute renal failure observed in patients that received either ibuprofen (5 or 10 mg kg−1) or paracetamol (12 mg kg−1). In a subset of 288 of the 795 infants or children that were hospitalized patients from the first study Lesko and Mitchell (1997) blood urea nitrogen (BUN) and creatinine levels were evaluated as markers for renal effects of the drugs. Mean BUN levels were ~4 mmol L−1 and creatinine levels were ~42 mmol L−1 in each of the three drug treatment groups. The prevelalence of BUN levels above 6.4 mmol L−1 and creatinine levels above 62 mmol L−1 was slightly higher in all the hospitalized patients where there was evidence of concomitant dehydration.

As noted earlier in the discussion of the large-scale paediatric study by Ashraf et al. (1999) there were no occurrences of renal failure or other serious renal conditions observed with either ibuprofen or paracetamol in the groups totalling 31,144 of younger or older children that were analysed.

Six cases of acute renal failure in children have been reported to be associated with ibuprofen and others with NSAIDs (Ulinksi et al. 2004). All patients recovered with normalization of serum creatinine levels after 308 days following cessation of the drug.

Thus, it would appear that although renal effects are known for ibuprofen as with the other NSAIDs there is a low risk of these adverse events occurring in children. No doubt dehydration (Leroy et al. 2007) and other factors can play an important role in the occurrence of renal effects form all NSAIDs in view of their concentrating in the renal tubular systems (Murray and Brater 1999).

Hepatotoxicity

Hepatic reactions have been of concern because of serious liver injury being reported with some NSAIDs and coxibs e.g. diclofenac, sulindac (in the USA), celecoxib, lumiracoxib (O’Brien and Bagby 1985; Stricker 1992; Cameron et al. 1996; Tolman 1998; Zimmerman 2000; Lacroix et al. 2004; Bannwarth and Berenbaum 2005; Chang and Schiano 2007) as well as paracetamol even at usual OTC doses (Watkins et al. 2006; Heard et al. 2007). The problem with attributing a particular drug whether it be an NSAID or otherwise, is that there are so many commonly used drugs that are hepatotoxic especially those drugs used by rheumatic patients e.g., antibiotiocs, anti-hypertensives, statins etc. (Stricker and 1992; Cameron et al. 1996; Zimmerman 2000). Moreover, the pattern of hepatopathies varies considerably among the different drugs may be taken with ibuprofen (or other NSAIDs) and that are associated with hepatoxicity (Table 32).

To obtain some indication of the incidence of hepatic reactions in patients taking NSAIDs Traversa et al. (2003) investigated the occurrence of hepatotoxicity in a cohort study in Umbria (Northern Italy) in subjects that had recently taken NSAIDs. A total of two events were recorded as “all hepatopathies” (of 122 cases that had NSAIDs) and two with liver injury associated (of 126 cases that had NSAIDs) with recent use of ibuprofen. Considering the extensive use of ibuprofen this is a low incidence.

In a “case/non-case” compilation of reports extracted from the FDA and WHO (NIMBUS) databases Sanchez-Matienzo et al. (2006) of the Pfizer Global Epidemiology group in Barcelona (Spain) attempted to give proportional estimates of the occurrence of different liver reactions attributed to individual NSAIDs. Tables 33 and 34 summarize data from Sanchez-Matienzo et al. Unfortunately, there are a number of critical issues about this data among them (a) there are no assessments of the likelihood of the event being associated with intake of a specific drug, (b) there is no information on confounding patient, disease or drug-related factors, (c) there is probably considerable double counting between the FDA and WHO data, (d) there is no information on the intake of drugs in DDD/100,000 patients, and (e) these data are in no sense quantitative and the WHO cautions especially on the use of the data from what are spontaneous reports. At best, these data only give signals. Thus, the data only show that ibuprofen has been reported to produce liver reactions with concomitant use of hepatotoxic drugs being implicated in a considerable proportion of cases.
Table 33

Proportion of reports (PRs) of various hepatic case definitions among cyclooxygenase (COX)-2 selective inhibitors and NSAIDs in the World Health Organization Uppsala Monitoring Centre data source, updated to the end of quarter 3 of 2003

Drug

Overall hepatic disorders

Abnormal hepatic function

Jaundice

Hepato-cellular damage

Non-infectious hepatitis

Hepatic failure

Total no. of reports

Rank

Bromfenac

20.7

10.8

3.2

3.5

4.3

2.2

2,057

14

Celecoxib

2.1

1.3

0.4

0.2

0.5

0.2

17,748

12

Diclofenac

4.7

3.2

1.0

0.2

1.4

0.1

21,082

5

Etodolac

3.6

2.5

1.0

0.4

1.2

0.3

3,553

9

Ibuprofen

1.8

1.1

0.4

0.2

0.5

0.1

32,973

13

Indomethacin

1.8

1.0

0.5

0.1

0.5

0.1

14,576

7

Ketorolac

0.6

0.4

0.1

0.1

0.1

0.2

1,867

6

Meloxicam

0.8

0.4

0.1

0.0

0.4

0.0

3,042

11

Multiple NSAIDsa

5.0

3.1

1.2

0.4

1.4

0.4

33,660

4

Naproxen

1.6

0.8

0.3

0.2

0.3

0.1

13,646

1

Nimesulide

14.4

7.2

2.0

1.0

5.7

0.4

1,057

3

Proxicam

2.0

1.2

0.4

0.1

0.6

0.1

13,973

8

Rofecoxib

1.5

0.8

0.2

0.1

0.4

0.1

20,429

2

Sulindac

9.9

5.2

3.2

0.5

3.1

0.2

5,777

10

From Sanchez-Matienzo et al. (2006). Values are percentages

PRs of concomitant use of other hepatotoxic drugs excluded

aIncludes reports involving > 1 COX-2 selective inhibitor and/or NSAID

Table 34

Frequencies of potential confounders in the US Food and Drug Administration (under Freedom of Information) (FDA/FOI) and World Health Organization Uppsala Monitoring Centre (WHO/UMC) data sources

Drug

Concomitant use of hepatotoxic drugs

Age ≥65 years

FDA/FOI

WHO/UMC

FDA/FOI

WHO/UMC

Nimesulide

46.8

13.2

27.8

20.3

Celecoxib

35.0

17.4

35.3

34.2

Sulindac

31.0

16.7

33.4

33.2

Meloxicam

30.1

5.9

38.0

13.5

Diclofenac

29.4

11.0

34.0

17.6

Etodolac

28.0

13.5

29.8

21.7

Indomethacin

27.9

17.1

25.0

18.0

Ibuprofen

25.0

18.1

17.6

13.8

Rofecoxib

20.9

5.8

38.4

21.4

Piroxicam

19.4

7.5

29.4

17.9

Naproxen

18.7

15.3

21.6

16.7

Ketorolac

17.6

69.6

22.0

19.1

Bromfenac

15.5

8.7

8.7

9.1

Multiple NSAIDsa

48.1

15.0

34.8

22.5

From Sanchez-Matienzo et al. (2006). Values are in percentages

aIncludes reports involving >1 cyclooxygenase-2 selective inhibitor and/or non-selective NSAID

Overall, the data suggest that hepatic reactions are probably rarely associated with ibuprofen. Since there have been no specific indications of reports of hepatic reactions with OTC use of ibuprofen from trials (Doyle et al. 1999; Kellstein et al. 1999; Boureau et al. 2004) or in literature analyses (Whelton 1995; Rainsford et al. 1997) it is likely that hepatotoxicity is not a significant risk factor at OTC dosages.

Hepatic reactions in children

Hepatic reactions do not appear to have been reported in any of the large scale hospitalization practitioner based studies in children (Lesko and Mitchell, 1995, 1999; Ashraf et al., 1999) or in critical reviews of clinical trails (Rainsford et al. 1997, 1999, 2001). Hepatitis has been frequently reported in trials of NSAIDs including ibuprofen and aspirin in JRA or JIA (Giannini et al. 1990) but in the small long-term study Ansell (1973) found that liver function tests were unaltered in these patients. The risks of liver reactions especially in JRA or JIA would appear to be low except where concomitant hepatotoxic drugs are taken (e.g. paracetamol, methotrexate) (Furst 1992; Hollingworth 1993).

Hypersensitivity reactions and asthma

NSAIDs are associated with the development of a range of hypersensitivity reactions including asthma. The symptoms of intolerance to these drugs ranges from severe bronchospasm that is often associated with nasal polyposis, rhino-conjunctivitis, urticaria, cervico-facial erythema, angio-oedema, hypotension and digestive disturbances (Arnaud 1995f). These symptoms may occur individually or in any combination (Arnaud 1995; de Weck et al. 2006). The symptoms may be manifest within a few minutes to several hours after ingestion of the drug.

These reactions have been frequently reported with aspirin and the terms “aspirin-associated asthma” or “aspirin-sensitive asthma” (ASA) have been employed to describe the association of symptoms of asthma with aspirin (Arnaud 1995; Rainsford 2004a: de Weck et al. 2006; Quiralte et al. 2007). In patients with ASA small doses of aspirin can lead to severe attacks. Frequently, sensitivity to NSAIDs overlaps that with aspirin, as well as with ibuprofen (Jenkins et al. 2004). This has given rise to the concept of sensitization and the use of desensitizing procedures to treat this condition (Rainsford 2004a, b; Jenkins et al. 2004; de Weck et al. 2006; Quiralte et al. 2007). Paracetamol has been found to be tolerated in some patients with NSAID intolerance but has in recent years been reported to be associated with asthma and hypersensivity reactions (Arnaud 1995).

The incidence of NSAID intolerance has been variously estimated depending on the method of evaluation, study design and population. Overall the incidence can be 0.6–2.5% of the general population (de Weck et al. 2006). The incidence is 4% when asthmatic patients are interviewed and can range up to 10–29% in adult patients with asthma or those that have been challenged (Arnaud 1995; de Weck et al. 2006). Hypersensivity to NSAIDs usualIy appears in the 2nd or 3rd decade and occurs in atopic subjects over the age of 40 years; this being more common in females than males (Arnaud 1995; de Weck et al. 2006). Aspirin or NSAID intolerance is occurs infrequently in children (de Weck et al. 2006). Aspirin does not normally involve sensitization through IgE (Arnaud 1995; de Weck et al. 2006).

The mechanisms of NSAID sensitivity have been debated over the years but there is some consensus that it is due to COX inhibition in susceptible individuals leading to over production of peptidoleukotries and accompanying symptoms of bronchoconstriction and other asthmatic symptoms (Arnaud et al. 1995; Rainsford 2005d). Other hypotheses suggest that there may be genetic influences relating to variability in leukotriene or prostanoid receptors (de Weck et al. 2006).

Asthma and hypersensitivity reactions in children

Asthma and hypersensitivity reactions have long been a cause for concern in children and the cross-reactions of ibuprofen with aspirin-sensitive asthma have been highlighted by several authors (Body and Potier 2004; Kidon et al. 2005; Mascia et al. 2005; Debley et al. 2005; Kanabar 2007; Ponvert and Scheinmann 2007). Two large scale studies in febrile asthmatic children (McIntyre and Hull 1996; Lesko et al. 2002) found that ibuprofen far from being associated with increased risk of asthma compared with paracetamol actually showed a slightly reduced risk. A randomised, double-blind, placebo-controlled, crossover bronchoprovocation challenge study in 100 prescreened school-aged children (6–18 years) found that ibuprofen induced bronchospasm was evidence in 2% of asthmatic subjects (Debley et al. 2005). Another 2% had clinically relevant decreases in spirometric measurements after ibuprofen administration but these did not meet the authors a priori criteria for a positive challenge test. These authors considered that ibuprofen-sensitive asthma has a low prevalence but none the less ibuprofen-induced bronchospasm should be considered as a risk in childhood asthma. A recent review also concludes that the risks of asthma are relatively low in children (Kanabar 2007).

Cutaneous reactions

Minor or “non-serious” skin reactions are among the more frequent reactions observed with NSAIDs including ibuprofen (Bigby and Stern 1985; O’Brien 1987; Ponvert and Scheinmann 2007). The risks of various skin reactions occurring with “ibuprofen containing medications” has been highlighted by Sanchez-Borges (2005). Their review mentioned the different types of skin reactions and the lack of quantitative information on the associations with ibuprofen.

A case–control study performed in Denmark of the occurrence of angio-oedema among NSAID and coxib users in hospital admissions by Downing et al. (2006) showed that the relative risks for this condition were higher in coxib users than those taking traditional NSAIDs. There were 25 cases out of a total of 377 patients.

Data of reports on serious and non-serious cutaneous reactions for NSAIDs reported in Italy as part of an overall programme of drug surveillance by Naldi et al. (1999) are shown in Fig. 10.
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig10_HTML.gif
Fig. 10

Reporting rates of serious (black bars) and non-serious cutaneous reactions to NSAIDs and analgesics compared with reports/consumption in DDD’s/1,000 inhabitants day−1 in 4 regions in Italy. Drug consumption data was derived from pharmacy sales data or hospital pharmacies. Numbers of reports for each drug are shown in brackets. Reproduced from Naldi et al. (1999), with permission

These show that ibuprofen ranked in the mid-range of reports. Thus, ibuprofen like other NSAIDs is associated with the occurrence of skin reactions many of which can be rated as mild. Serious ADRs in the skin are rare. There have been occasional reports of Stevens-Johnson and Lyell’s syndromes as well as severe bullous reactions (Bigby and Stern 1985; Miwa and Jones 1999; Sanchez Borges et al. 2005). However, these serious conditions have not been reported in controlled trials or literature on OTC events from ibuprofen (Rainsford et al. 1997, 2001; Doyle et al. 1999; Kellstein et al., 1999; Hersh et al. 2000a; Ashraf et al. 2001; Le Parc et al. 2002; Boureau et al. 2004).

Skin reactions in children

In the large scale Boston University Fever studies Lesko and Mitchell (1995, 1999) did not observe any hospitalizations for anaphylaxis. However, three cases of erythema multiforme occurred in patients that had received ibuprofen and one that had received paracetamol, thus suggesting that the risks of these events very low (Lesko and Mitchell, 1995). Ashraf et al. (1999) in their large scale trial also noted there were no cases of Stevens-Johnson syndrome among their patients.

Other adverse reactions

Rare adverse events that have been reported at prescription-level doses with ibuprofen and less frequently with OTC doses are common to those seen with all NSAIDs. Among these thrombocytopaenia, agranulocytosis, anaemia) aseptic meningitis and anaphylactoid reactions, interactions with the immune, endocrine and metabolic systems, central nervous system and ocular effects (Hoffman and Gray 1982; O’Brien and Bagby 1985; Haupt et al. 1991; Miwa and Jones 1999; Jackson et al. 2006; Layton et al. 2006; Neuman and Nicar 2007). Most but not all these adverse events are rare with the exception of allergies including aspirin-sensitive asthma, especially with OTC dosages of ibuprofen.

The possibility of fractures being associated with NSAID use has been identified in patients with rheumatoid and osteoarthritis (Vestergaard et al. 2006). In these epidemiological studies adjustments were made for stratifying two cumulative daily dosages (defined daily dose DDD) and other confounders. There was an odd disease association that was observed in these studies in that osteoarthritis was associated with a decreased risk of any fracture and rheumatoid arthritis was associated with increased development of fractures. These studies highlighted that high dosage intake of aspirin, paracetamol, diclofenac, meloxicam and some other NSAIDs but not coxibs was associated with an increased risk of fractures. Ibuprofen showed an odd inverse dose-related effect in as much as the adjusted odds ratios (less than or 20–74 DDDs approx 1.8–1.82) were higher than those where the drug was taken in greater quantities (1.42). Thus, hip fractures and other bone fractures are a risk factor in elderly rheumatic patients taking NSAIDs for long periods of time maybe more a class indication as distinct from a specific risk factor associated (i.e. common to NSAIDs as a group) with anyone drug.

Using the UK general practice data base Van Staa et al. (2000) examined the factor risk exposed to NSAIDs using a case control approach. Regular NSAID intake was associated with an increase of risk compared with control of 1.47 (1.42–1.52 95% CI) of non-vertebral fractures while the risk of hip fractures was relatively low being 1.08 (0.9–1.19 95% CI) it appeared from this study that ibuprofen had the lowest risk of non-vertebral fractures as it was used as the reference but there was a larger number of cases of ibuprofen compared with other NSAIDs probably reflecting a wider-spread use. In contrast to these observations on the development of fractures studies by Persson et al. (2005) suggest that long-term treatment of patients who have undergone revision of hip arthroplasty.

In children some of the abovementioned rare ADRs have not been reported. During the 1990s there was concern about the possibility of ibuprofen being associated with the development of necrotizing fasciitis in previously healthy children who had primary varicella (Khan and Styrt 1997; Neutel and Pless 1997). This was reinforced by a case–control study that hospitalizations for this condition had an OR of 11.5 (95% CI 1.4–96.6) (Zerr et al. 1999). The high range of CI values makes this study of dubious value for ascertainment of risks.

Lesko et al. (2001) undertook study involving 25 tertiary care hospitals from across the USA in which a total of 97 children were identified with possible group A streptococcal infections with primary varicella infections. They found that the risk of necrotizing fasciitis was not associated with use of ibuprofen. Moreover, in their large scale studies Lesko and Mitchell (1995, 1999) and Ashraf et al. (1999) have failed to find evidence of ibuprofen being associated with necrotozing fasciitis.

Hypothermia is rarely reported to be associated with ibuprofen but other NSAIDs are also associated with this AE (Desai and Sriskandan 2003).

Cases of poisoning in UK

The numbers of reports to the National Poisons Advisory Service (NPIS; the main referral centre in the UK for enquiries and reports of poisonings) those from ibuprofen are now second to those of paracetamol (Volans and Fitzpatrick 1999). Cases of poisoning from ibuprofen considered by Volans and Fitzpatrick (1999) were reviewed principally in the period when the drug was available prescription in the UK. It is clear from a large number of studies (Rainsford 1999c, 2004a) that ibuprofen has the lowest toxicity of all analgesics. Mechanistically, there is no evidence of irreversible toxic actions that could be attributed to the covalent modification of endogenous biomolecules analogous to that observed in the liver toxicity from quinine–imine metabolite of paracetamol (Graham and Hicks 2004) or the GI ulceration and bleeding arising from irreversible acetylation of platelets and other cyclo-oxygenases or biomolecules from the acetyl-moiety of acetylsalicylic acid (aspirin) (Rainsford et al. 1981, 1984; Rainsford 2004a; Graham et al. 2004). More recently, the assessment by Moore (2007) of the risks of serious events from overdose has concluded that ibuprofen is essentially benign.

Impact of limitations on analgesic pack sizes in UK

During the 1990s, there was growing concern in the UK about the toxicity to the liver from paracetamol. This drug is known to cause irreversible liver damage often with fatal outcome in poisoning. There is also evidence that this drug may cause severe liver injury in certain conditions in doses ≥4 g day−1.

As a consequence of these concerns the UK Medicines and Healthcare Products Authority (MHRA) reviewed the safety of all analgesics since there have been ongoing concerns about gastro-intestinal and renal effects of aspirin, in particular, and to a lesser extent from ibuprofen.

A consequence of this was that legislation was introduced in September 1998 to limit the general sales and sales by pharmacies of paracetamol and aspirin by restricting the pack size (Morgan et al. 2007). Ibuprofen with or without codeine had already been sold in small pack sizes consistent with its OTC labelling. The limitation of pack sizes of paracetamol or aspirin has had a significant impact in reducing cases of poisoning, accidental or deliberate, with a trend in reducing in serious outcomes including deaths (Hawton et al. 2001; Morgan et al. 2007).

Recent statistics from the UK National Poisonings Information Service (Anonymous 2006) show that over 115,000 enquiries have been received about paracetamol annually that comprised 99,000 visits to paracetamol poisoning information on TOXBASE—the NPIS’s online information database—and about 16,000 telephone calls. In comparison 42,000 enquiries were received by the NPIS about ibuprofen and 25,000 about aspirin.

General safety assessment

Assessment of safety in children (0.3–12 year old)

The question of assessing the safety of NSAIDs in the group of 0.3–-12 year old children has been the subject of debate as to whether studies in “Adults” aged >18 years and the “Elderly” variously aged ≤65 years can be extrapolated to children and is so upon what criteria? It is arguable that hormonal and developmental influences can make for marked variations in the enzymes and receptors involved in the expression and control of inflammatory states thus influencing the pharmacodynamics and toxicological potential of NSAIDs. Furthermore, variations in liver and extra-hepatic metabolism and elimination of NSAIDs could affect their pharmacokinetics in young adults compared with that in adults. Given these theoretical issues it is of relevance to ask how this relates to ibuprofen having appreciable variations in PK and PD in children >1–2 years (Table 3) c.f. adults (Tables 1, 2) and the degree to which in the absence of information on these parameters that studies in adults can be extrapolated from the extensive data in adult populations.

First, it is clear that dose-adjustment for age in the children and infants is a basic factor which as long as the recommended dosages of ibuprofen are taken effectively controls drug exposure in this age group. Second, it is clear that the pharmacokinetics of ibuprofen and especially the drug metabolizing systems differ little if at all with age in children c.f. adults (Tables 1 c.f. 2, 3).

From the point of view of risks for children who receive the oral formulation of the drug at what is in general much lower doses than the prescription level in adults or in severe conditions (e.g. arthritis) it is reasonable to conclude that risks from exposure to ibuprofen especially when it is largely given for short periods of time in moderate to low doses for acute conditions that the occurrence of serious ADRs would be relatively low and acceptable from a health and safety viewpoint. The evidence of relatively low occurrence of serious ADRs or rare events from large scale practitioner or centre based studies (Lesko and Mitchell, 1995, 1999; Lesko et al., 2001, 2002; Ashraf, 1999) and critical reviews of clinical trials (Rainsford et al. 1997, 1999, 2001). The relatively low intrinsic toxicity of the drug also gives support to the view that it has a high degree of tolerance from the point of view of human toxicity.

Overall safety assessment

Overall, the data reviewed here shows that in comparison with other NSAIDs it appears that ibuprofen has relatively low risks for GI and CV ADRs, especially in the more serious category. Some attempts to rate risks of ADRs from NSAIDs against known life events has recently been considered by Moore et al. (2008). They point out that people tend to underestimate common risks and overestimate rare risks. Also people primarily respond emotionally to risk and are risk-averse when confronted with medical interventions. Taking these and other factors into account these authors have devised a rating of (amongst other ADRs) serious GI and CV risks against known life events (Fig. 11). The risks of these two serious events occurring with ibuprofen is low although as the graph shows the points for each ADR with a scattered group it is hard to distinguish differences among the NSAIDs. It should be pointed out that there risks are from the large scale trials with prescription levels NSAIDs performed long-term. Thus, these estimates represent risks at what could be regarded as the “high end” of drug exposure.
https://static-content.springer.com/image/art%3A10.1007%2Fs10787-009-0016-x/MediaObjects/10787_2009_16_Fig11_HTML.gif
Fig. 11

Risks of GI bleed or heart attack from NSAIDs. From Moore et al. (2008), reproduced with permission. • = GI risks. ▲ = risks of heart attack. Reproduced on basis of Creative Commons Attribution Licence (http://creativecommons.org/licenses/by/2.0)

While there have only been limited investigations into the effects of application of the suppository formulation in children these and the ADR data show that it is an acceptable formulation that can be applied to children. Further preclinical and clinical investigations, as well as epidemiological studies should be undertaken to establish further the safety and efficacy of ibuprofen suppositories.

In conclusion, there is substantial evidence of safety and efficacy in support of the view that ibuprofen is a safe and very effective drug for use in a range of painful, acute and chronic inflammatory conditions.

Copyright information

© Birkhäuser Verlag, Basel/Switzerland 2009