The FASEB Journal • Research Communication
Acetaminophen (paracetamol) is a selective cyclooxygenase-2 inhibitor in man Burkhard Hinz,*,1 Olga Cheremina,† and Kay Brune† *Institute of Toxicology and Pharmacology, University of Rostock, Rostock, Germany; and †Institute of Experimental and Clinical Pharmacology and Toxicology, Friedrich Alexander University Erlangen-Nu¨rnberg, Erlangen, Germany For more than three decades, acetaminophen (INN, paracetamol) has been claimed to be devoid of significant inhibition of peripheral prostanoids. Meanwhile, attempts to explain its action by inhibition of a central cyclooxygenase (COX)-3 have been rejected. The fact that acetaminophen acts functionally as a selective COX-2 inhibitor led us to investigate the hypothesis of whether it works via preferential COX-2 blockade. Ex vivo COX inhibition and pharmacokinetics of acetaminophen were assessed in 5 volunteers receiving single 1000 mg doses orally. Coagulation-induced thromboxane B2 and lipopolysaccharideinduced prostaglandin E2 were measured ex vivo and in vitro in human whole blood as indices of COX-1 and COX-2 activity. In vitro, acetaminophen elicited a 4.4fold selectivity toward COX-2 inhibition (IC50ⴝ113.7 mol/L for COX-1; IC50ⴝ25.8 mol/L for COX-2). Following oral administration of the drug, maximal ex vivo inhibitions were 56% (COX-1) and 83% (COX-2). Acetaminophen plasma concentrations remained above the in vitro IC50 for COX-2 for at least 5 h postadministration. Ex vivo IC50 values (COX-1: 105.2 mol/L; COX-2: 26.3 mol/L) of acetaminophen compared favorably with its in vitro IC50 values. In contrast to previous concepts, acetaminophen inhibited COX-2 by more than 80%, i.e., to a degree comparable to nonsteroidal antiinflammatory drugs (NSAIDs) and selective COX-2 inhibitors. However, a >95% COX-1 blockade relevant for suppression of platelet function was not achieved. Our data may explain acetaminophen’s analgesic and antiinflammatory action as well as its superior overall gastrointestinal safety profile compared with NSAIDs. In view of its substantial COX-2 inhibition, recently defined cardiovascular warnings for use of COX-2 inhibitors should also be considered for acetaminophen.—Hinz, B., Cheremina, O., Brune, K. Acetaminophen (paracetamol) is a selective cyclooxygenase-2 inhibitor in man. FASEB J. 22, 383–390 (2007) ABSTRACT
Key Words: cyclooxygenase isoenzymes 䡠 COX-2 selectivity 䡠 human whole blood assay 䡠 pharmacokinetics Acetaminophen (INN, paracetamol) is one of the most widely used over-the-counter antipyretic and analgesic drugs worldwide. It is recommended as first-line therapy for pain associated with osteoarthrosis (1). 0892-6638/08/0022-0383 © FASEB
Although discovered more than 100 years ago and used extensively for ⬃50 years, its mode of action is still unclear. The analgesic and antipyretic actions of acetaminophen resemble those of nonsteroidal antiinflammatory drugs (NSAIDs). However, it is commonly stated that acetaminophen acts centrally and is at best a weak inhibitor of prostaglandin (PG) synthesis by cyclooxygenase (COX) -1 and COX-2 (2). This concept is based on early work by Flower and Vane (3), who showed that PG production in brain is 10 times more sensitive to inhibition by acetaminophen than that in spleen. However, this finding was not supported by later studies (4 – 6). Moreover, it turned out that acetaminophen elicits no measurable inhibition of PG formation in broken cell preparations but a profound suppression in intact cells (7). Attempts to explain the pharmacological action of acetaminophen as inhibition of a central COX isoform, derived from the same gene as COX-1 and referred to as COX-3 (8), have been meanwhile rejected for several reasons (9). In particular, the existence of a functional human COX-3 has been questioned, given that retention of intron 1 in human COX-3 leads to a shift in the reading frame, a premature termination, and a truncated, COX-inactive protein (10, 11). Qin et al. (12) reported a low-level expression of three splice variants of COX-1 in human tissues but were not able to show a significantly different potency of acetaminophen in inhibiting human COX-1 vs. an intron 1-retained COX-1 splice variant. Thus, research for the real target of acetaminophen proceeds. The pharmacological profile of acetaminophen is very similar to that of selective COX-2 inhibitors (coxibs). As coxibs, acetaminophen given orally at recommended single doses elicits no toxic effect on the gastrointestinal tract (13), does not inhibit platelet function (14, 15), and hardly provocates bronchoconstriction in aspirin-sensitive asthmatics (16). Consequently, the fact that acetaminophen is acting functionally as a coxib led us to investigate the hypothesis 1
Correspondence: Institute of Toxicology and Pharmacology, University of Rostock, Schillingallee 70, D-18057 Rostock, Germany. E-mail:
[email protected] doi: 10.1096/fj.07-8506com 383
whether acetaminophen works via selective COX-2 inhibition. We here show that oral administration of 1000 mg acetaminophen to human volunteers inhibits blood monocyte cyclooxygenase (COX)-2 by more than 80%, i.e., to a comparable degree as NSAIDs and selective COX-2 inhibitors. By contrast, a COX-1 blockade relevant for inhibition of platelet function was not achieved. Our data may explain acetaminophen’s favorable gastrointestinal safety profile as well as its analgesic and antiinflammatory action. Moreover, in view of its substantial COX-2 inhibition, a re-evaluation of acetaminophen’s cardiovascular risks appears to be warranted. MATERIALS AND METHODS In vitro investigations Materials Acetaminophen, aspirin, and lipopolysaccharide (LPS) from Escherichia coli (serotype 026:B6) were obtained from Sigma Chemie (Steinheim, Germany). PGE2 and TxB2 enzyme immunoassay kits were from Cayman (Ann Arbor, MI, USA). Effects of acetaminophen on COX-1 and COX-2 activity in human whole blood COX-1 assay Blood was drawn from healthy volunteers who had not taken any NSAID 2 wk prior to blood sampling. Aliquots of whole blood without anticoagulant were immediately transferred to glass tubes containing test agent or vehicle. Blood was allowed to clot for 1 h at 37°C (16). Serum was separated by centrifugation, and serum TxB2 levels were determined. COX-2 assay Aliquots of heparinized whole blood from healthy volunteers were incubated with LPS (10 g/ml) plus test agent or vehicle for 24 h at 37°C (17). The contribution of platelet COX-1 activity was inhibited by the addition of aspirin (10 g/ml) at the start of the incubation. Plasma was separated by centrifugation, and PGE2 levels were determined subsequently. For both enzymatic assays, concentration response curves were fitted by a sigmoidal regression with variable slope, and 50% inhibitory concentration (IC50) values were derived by use of PRISM® Version 3.0 (GraphPad, San Diego, CA, USA). The degree of COX-1 or COX-2 inhibition was calculated as the percentage change of plasma eicosanoid (COX-1: TxB2; COX-2: PGE2) in the presence of acetaminophen. In the case of COX-2, from each PGE2 value in the LPS- and LPS/ acetaminophen-treated groups, basal PGE2 levels were substracted before calculating percentage change. Effects of acetaminophen on COX-2 activity in short-term assays Human recombinant COX-2 The inhibitory effect of acetaminophen on the human recombinant COX-2 enzyme was determined using the COX Inhib384
Vol. 22
February 2008
itor Screening Assay from Cayman. Experiments were performed according to the manufacturer’s instruction using 12-min incubation periods of the active enzyme with acetaminophen. In brief, the assay directly measures PGF2␣ produced by tin chloride reduction of COX-2-derived PGH2. COX-2 from human blood monocytes Mononuclear cells were isolated from heparinized human whole blood of healthy volunteers by density gradient centrifugation with Histopaque-1077 as described previously (18). Cells seeded in 48-well culture plates at a density of 1 ⫻ 106 cells per well were allowed to adhere for 3 h. After removal of nonadherent cells by washing, adherent monocytes were cultured in RPMI 1640 medium. Incubations were performed under serum-free conditions in a humidified incubator at 37°C and 5% CO2. To assess the effect of test compounds on COX-2 activity, monocytes were treated with aspirin (250 mol/L) for 2.5 h to inactivate endogenous COX activity. Thereafter, cells were washed extensively and subsequently incubated with vehicle (group 1) or 10 g/ml LPS (groups 2 and 3) for 18 h to induce COX-2. Following extensive washing and medium change, vehicle (groups 1 and 2) or 100 mol/L acetaminophen (group 3) was added to the cultures followed by a 90-min incubation period. Afterward, the cell culture supernatants were removed and analyzed for PGE2. The degree of COX-2 inhibition was calculated as the percentage change of PGE2 levels in LPS/acetaminophen-treated cells relative to cells treated with LPS only. To this end, basal PGE2 levels (group 1) were substracted from each value obtained from groups 2 and 3. Effects of acetaminophen on COX-2 expression Western blot analysis of human blood monocytes Mononuclear cells were seeded in 6-well culture plates at a density of 5 ⫻ 106 cells per well and were allowed to adhere for 3 h. After removal of nonadherent cells by washing, adherent monocytes were treated for 24 h with either vehicle, acetaminophen (100 mol/L), LPS (10 g/ml), or LPS plus acetaminophen. Afterward, cells were washed, harvested, and pelleted by centrifugation. Cells were than lysed in solubilization buffer (50 mM HEPES, pH 7.4; 150 mM NaCl; 1 mM EDTA; 1% (v/v) Triton® X-100; 10% (v/v) glycerol; 1 mM phenylmethylsulfonyl fluoride; 1 g/ml leupeptin; and 10 g/ml aprotinin). After incubation on ice for 30 min, the lysates were centrifuged at 10,000 g for 5 min. Supernatants were used for Western blot analysis. Proteins were separated on a 10% sodium dodecyl sulfate-polyacrylamide gel. Following transfer to nitrocellulose and blocking of the membranes with 5% milk powder, blots were probed with a specific antibody raised to COX-2 (BD Biosciences, Heidelberg, Germany) or -actin (Calbiochem, Bad Soden, Germany), the latter being used as a loading control. Subsequently, membranes were probed with horseradish peroxidase-conjugated Fab-specific anti-mouse IgG (Sigma, Steinheim, Germany). Antibody binding was visualized by enhanced chemiluminescence Western blotting detection reagents (Amersham Pharmacia Biotech, Freiburg, Germany). Densitometric analysis of COX-2 band intensities (normalized to -actin) was achieved by an optical scanner (Gel Doc 2000) and the Multi-Analyst program, version 1.1 (both from Bio-Rad Laboratories, Hercules, CA, USA).
The FASEB Journal
HINZ ET AL.
Pharmacokinetics of acetaminophen and ex vivo inhibition of COX activities after acetaminophen treatment Subjects and study design Five male volunteers (all of them doctors of medicine), 39 to 65 years (mean age: 44.8 years) with a mean weight of 69.0 ⫾ 2.9 kg (mean⫾se), participated in the study. Subjects did not take any other medication (including aspirin or other NSAIDs) within 2 wk before and throughout the study. Volunteers took 1000 mg acetaminophen (2 tablets Paracetamol-ratiopharm® 500, ratiopharm GmbH, Ulm, Germany). Acetaminophen was administered between 8:00 and 8:30 a.m. after an overnight fast. For pharmacokinetic studies and COX activity assays, peripheral venous blood samples were taken from each volunteer immediately before and at 0.25, 0.75, 1.5, 3, 5, 8, 10, and 24 h after administration of acetaminophen. For determination of acetaminophen, heparinized blood samples were centrifuged and plasma aliquots were frozen. Until further analysis, plasma samples were stored at –20°C for a maximum of 1 wk.
the absence of LPS were substracted from PGE2 levels determined in LPS-treated blood aliquots. Maximal observed inhibition of the COX isoforms and times to reach it were obtained directly from the effect vs. time curves. The areas within the effect-time curves (AWECs) were calculated using the linear trapezoidal rule. Larger AWECs correspond to greater levels of COX inhibition. Pharmacokinetic analysis Plasma concentration-time curves of acetaminophen were evaluated by noncompartmental analysis using WinNonlin® Version 3.3 (Pharsight, Mountain View, CA, USA). Maximal plasma concentrations (Cmax) and times to Cmax (tmax) were obtained directly from the individual plasma concentration vs. time curves. The area under the plasma concentrationtime curve up to the last quantifiable plasma concentration (AUCt) was determined according to the linear trapezoidal method. Correlation between pharmacodynamics and pharmacokinetics
Ex vivo inhibition of COX activities COX-1 assay Immediately after blood sampling, whole blood samples without anticoagulant were incubated for 1 h at 37°C and subsequently centrifuged, and serum TxB2 levels were determined.
For assessing the correlation between plasma concentrations of acetaminophen and changes in COX-1 or COX-2 inhibition, plasma concentration-response curves were fitted using a sigmoidal regression with variable slope and the ex vivo IC50 values for COX-1 or COX-2 inhibition were derived by using PRISM® Version 3.0 (GraphPad).
COX-2 assay
RESULTS
Immediately after blood sampling, heparinized whole blood samples were incubated with 10 g/ml LPS for 24 h at 37°C. The contribution of platelet COX-1 activity was inhibited by the addition of aspirin (10 g/ml) at the start of the incubation. Plasma was separated by centrifugation, and PGE2 levels were determined.
In vitro effects of acetaminophen on COX-1 and COX-2 activity in human whole blood
Determination of acetaminophen in human plasma Acetaminophen was analyzed by high-performance liquid chromatography (HPLC) with ultraviolet (UV) detection following a previously published method (19). In brief, calibration standards and samples were prepared by adding 0.2 ml 10% aqueous perchloric acid to 0.2 ml plasma, vortexed for 10 s, and centrifuged for 5 min at 4000 r.p.m. Clear supernatants (50 l each) were injected directly into the HPLC system with a reversed-phase column (CC 125/4 Nucleosil 120 –3 C18; Machery-Nagel, Du¨ren, Germany) and a CC 8/4 Nucleosil C18 precolumn insert. The mobile phase consisted of a 96:4 (v/v) mixture of 20 mM sodium dihydrogen phosphate dihydrate (pH⫽2.5) and acetonitrile. The flow rate was 0.9 ml/min. UV detection was set at 254 nm. Acetaminophen was eluted at 7.0 min. Acetaminophen concentrations were calculated as equivalents from the peak area values. Linearity of the calibration curves was proven from 0.2 to 40 g/ml with regression coefficients 0.999. Pharmacodynamic analysis The degree of COX-1 or COX-2 inhibition was calculated as the percentage change of plasma eicosanoid (COX-1: TxB2; COX-2: PGE2) measured at different time points postadministration relative to predose plasma eicosanoid levels. In the case of COX-2, for each value, basal PGE2 levels measured in SELECTIVE COX-2 INHIBITION BY ACETAMINOPHEN
Using in vitro whole blood approaches to determine COX inhibitory profiles described above, acetaminophen was shown to elicit a concentration-dependent inhibition of either isoform (Table 1A). On the basis of the estimated IC50 values, acetaminophen displayed a 4.4-fold selectivity toward COX-2. Pharmacokinetics of acetaminophen The aim of the in vivo approach was to determine whether acetaminophen plasma concentrations relevant to COX inhibition are reached in volunteers treated with the drug at a recommended single oral dose of 1000 mg. Average plasma concentrations of acetaminophen remained below the in vitro IC50 value for COX-1 inhibition but were greater than or equal to the in vitro IC50 value for COX-2 inhibition for at least 5 h after administration (Fig. 1). Pharmacokinetic data of acetaminophen are summarized in Table 1B. Ex vivo inhibition of COX activity Time-courses of ex vivo LPS-induced PGE2 levels and coagulation-induced TxB2 levels in blood from acetaminophen-treated volunteers revealed a profound inhibition of COX-2 with a maximal inhibition of 83% and a mean inhibition of 66% during the first 385
TABLE 1. Apparent in vitro potency of acetaminophen as an inhibitor of COX-1 and COX-2 activity and the pharmacokinetic and pharmacodynamic parameters of acetaminophen after oral administration of a 1000-mg dose Parameter
Value
A. In vitro COX inhibition COX-1 IC50 (mol/L) COX-2 IC50 (mol/L) Selectivity ratio (COX-1/COX-2) B. Pharmacokinetics tmax (h) Cmax (mol/L) Fold increase of Cmax over COX-1 IC50 Fold increase of Cmax over COX-2 IC50 AUCt (mol · h/L) t1/2 (h) C. Ex vivo COX-1 inhibition Time to maximal inhibition (h) Maximal inhibition (%) Mean inhibition0.25–5 h (%) Mean inhibition0.25–24 h (%) AWEC0–5 h (% · h) AWEC0–24 h (% · h) IC50 (mol/L) D. Ex vivo COX-2 inhibition Time to maximal inhibition (h) Maximal inhibition (%) Mean inhibition0.25–5 h (%) Mean inhibition0.25–24 h (%) AWEC0–5 h (% · h) AWEC0–24 h (% · h) IC50 (mol/L)
113.7 ⫾ 17.2 25.8 ⫾ 1.8 4.41 1.25 ⫾ 0.48 104.8 ⫾ 24.0 0.92 4.06 416.1 ⫾ 29.0 4.47 ⫾ 0.58 0.95 ⫾ 0.24 55.8 ⫾ 4.9 37.4 ⫾ 4.4 29.8 ⫾ 4.7 181.9 ⫾ 16.8 500.3 ⫾ 123.7 105.2 1.65 ⫾ 0.86 83.2 ⫾ 6.5 66.2 ⫾ 4.7 47.5 ⫾ 2.1 323.9 ⫾ 16.5 640.8 ⫾ 81.9 26.3
Figure 1. Plasma concentrations of acetaminophen following oral administration of 1000 mg. Values are means ⫾ se from n ⫽ 5 volunteers. At 24 h postadministration, plasma concentrations of acetaminophen were only detectable in 3 of 5 volunteers. Superimposed horizontal lines represent IC50 values of acetaminophen for COX-1 and COX-2 inhibition obtained in vitro (Table 1A).
In vitro effects of acetaminophen on COX-2 activity in short-term assays and on COX-2 expression in human blood monocytes Further experiments were performed to determine whether the suppression of COX-2-dependent PGE2 generation by acetaminophen observed in long-term whole blood assays was a result of direct inhibition of COX-2 enzyme activity or decreased COX-2 expression.
Values in A are mean ⫾ se from 4 blood donors. Except IC50 values and fold increases, all data in B–D are expressed as mean ⫾ se from 5 healthy male volunteers that took 1000 mg acetaminophen orally. IC50 , 50% inhibitory concentration; Cmax, maximal plasma concentration; tmax, time to Cmax; AUCt, area under the plasma concentration-time curve from time 0 to the time of the last measurable concentration; t1/2 , terminal half life; AWEC, area within the effect-time curve.
5 h postadministration (Table 1D and Fig. 2). COX-1 was inhibited only by up to 56% (Fig. 2 and Table 1C). Times to maximal inhibition of COX-1 and COX-2 were in accordance with the time-to-maximal acetaminophen plasma levels (comp. Table 1B–D). Correlation between acetaminophen plasma concentrations and ex vivo inhibition of COX isoforms The relationship of acetaminophen plasma concentrations to ex vivo inhibition of COX-1 and COX-2 was examined graphically and explored by estimating the acetaminophen plasma concentration required to produce 50% inhibition of the respective COX isoform. The calculated ex vivo IC50 values (Fig. 3) compared favorably with the in vitro IC50 values of acetaminophen for inhibition of both enzymes (Table 1A). 386
Vol. 22
February 2008
Figure 2. Time-dependent ex vivo inhibition of COX-1 (A) and COX-2 activity (B) following oral administration of 1000 mg acetaminophen to 5 volunteers. Values are means ⫾ se.
The FASEB Journal
HINZ ET AL.
DISCUSSION The mode of action of acetaminophen still belongs to the pharmacological enigmas in pain research. Whereas the drug was first associated with a predominant suppression of PG synthesis in brain tissue (3), subsequent studies contradicted a cell-specific modulation (4 –7). One of the reasons leading to such controversial findings may lie in the fact that acetaminophen elicits no measurable inhibition of PG formation in broken cell preparations but a profound suppression in intact cells (7). Despite these findings, acetaminophen is still regarded as a weak inhibitor of peripheral COX-1 and COX-2 enzymes (2). However, evidence is emerging to suggest that acetaminophen shares many properties with selective COX-2 inhibitors in that both elicit
Figure 3. Relationship between ex vivo inhibition of COX-1 (A) and COX-2 activity (B) and plasma concentrations of acetaminophen. For analysis of ex vivo IC50 values, plasma drug concentrations and corresponding COX inhibitions determined after single-dose administration of 1000 mg acetaminophen were included. Values were derived from n ⫽ 5 volunteers.
In a first experimental approach the impact of a short-term treatment with acetaminophen on COX-2 activity was investigated using the human recombinant COX-2 enzyme as well as monocytes with preinduced COX-2. According to Fig. 4A, acetaminophen at a therapeutic concentration of 100 mol/L elicited a substantial inhibition of COX-2 enzyme activity in both systems. COX-2 activity values obtained in human whole blood treated with the same concentration of acetaminophen (data are part of the experiments in Table 1A) were included in Fig. 4A for comparison. In a second approach we addressed the question of whether the inhibition of COX-2 by acetaminophen in the whole blood assays that lasted 24 h was a consequence of decreased COX-2 expression. To this end, the impact of acetaminophen on COX-2 protein levels both in the presence and absence of LPS was analyzed in Western blotting experiments. According to Fig. 4B, acetaminophen at 100 mol/L left basal as well as LPS-stimulated COX-2 levels virtually unaltered. SELECTIVE COX-2 INHIBITION BY ACETAMINOPHEN
Figure 4. Effect of acetaminophen (100 mol/L) on COX-2 activity of the human recombinant COX-2 enzyme, human blood monocytes, and human whole blood (A) and impact of acetaminophen (100 mol/L) on COX-2 protein levels in human blood monocytes (B). COX-2 activity (A) and COX-2 protein (B) were determined according to the protocols described under Material and Methods. Percent control (A, B) represents comparison with vehicle-treated cells (100%) in the absence of test substance. Values in (A) are means ⫾ se of n ⫽ 3– 6 experiments. COX-2 protein levels (B) are shown by one representative Western blot. -actin was used as a loading control. Percentual values above the blots are means ⫾ se of n ⫽ 4 blots. 387
no toxic effect on the gastrointestinal tract (13), do not inhibit platelet function (14, 15), and hardly provocate bronchoconstriction in aspirin-sensitive asthmatics (16). On the basis of these clinical observations, we investigated the hypothesis of a selective COX-2 inhibition by acetaminophen using the human whole blood assay (17). The latter is considered to provide the most meaningful index of selectivity by measuring COX inhibition in a physiological milieu, thereby ensuring intact cellular interactions influencing arachidonic acid metabolism and drug binding to plasma proteins. The present study demonstrates that acetaminophen displays an ⬃4-fold selectivity for inhibition of COX-2 both in vitro and in vivo. Administration of a standard dose of the drug caused an almost complete inhibition of COX-2 in human volunteers, whereas only a moderate inhibition of COX-1 was observed. In our hands, the inhibition of COX-2 activity by acetaminophen was even higher than that by single-dose administration of celecoxib and rofecoxib at clinically recommended doses of 200 and 25 mg, respectively (20). In contrast to the latter compounds, no substantial between-patient variability in the plasma concentrations of acetaminophen and the corresponding degree of ex vivo COX-2 inhibition was observed. Though not investigated, it is obvious from the present data that the recommended 4-times-daily dosage regime of 1000 mg acetaminophen is associated with a permanent 60 – 80% inhibition of COX-2. Previous data (15) suggesting a much lower and unselective ex vivo inhibition of COX-2 by acetaminophen are likely due to the analysis of a limited timeframe. Considering the similar behavior of acetaminophen and coxibs on COX-2-derived PGE2 levels in human whole blood, our data indicate that acetaminophen could influence peripheral COX-2 during inflammation similar to rofecoxib and celecoxib. Although such assumption appears justifiable, acetaminophen has been claimed to possess no antiinflammatory activity. In reality, this statement is based largely on early work showing that acetaminophen does not suppress inflammation associated with rheumatoid arthritis (21, 22). This might be explained by the high extracellular concentrations of arachidonic acid and peroxide in the inflamed tissue, both of which diminish the effect of acetaminophen on PG synthesis (7, 23, 24). On the other side, acetaminophen decreases tissue swelling following oral surgery in humans, with activity very similar to that of ibuprofen (25, 26). A peripheral antiinflammatory action is further supported by findings showing acetaminophen as an inhibitor of nociception and edema in the rat carrageenan footpad model (27, 28), an inflammatory condition critically dependent on COX-2-derived PGs (29). Our results further indicate that the COX-2 blockade by acetaminophen in the human vasculature should cause concerns. In fact, a permanent blockade of COX-2-dependent PGs including prostacyclin and PGE2 is the currently most plausible explanation for the cardiovascular hazard conferred by selective and 388
Vol. 22
February 2008
nonselective COX-2 inhibitors (20, 30). In this context, long-term increase of blood pressure has been proposed to underlie cardiovascular side effects occurring after prolonged use of these compounds. Regarding the effects of acetaminophen on blood pressure, a prospective cohort study found that regular consumption of acetaminophen was associated with a significantly higher risk for development of hypertension compared with no use (31). Noteworthy, the relative risk of acetaminophen was similar to that of NSAIDs. In line with this notion, a recently published large, prospective study showed that use of acetaminophen at more than 15 tablets per week confers nearly the same risk for cardiovascular events as traditional NSAIDs (32). In our hands acetaminophen did not decrease COX-2 protein levels in human blood monocytes both in the absence and presence of LPS, thus excluding a significant contribution of impaired COX-2 expression to acetaminophen’s inhibitory action on COX-2-derived PGE2 levels in human blood. Moreover, inhibition of COX-2 activity by acetaminophen was confirmed in short-term experiments using the human recombinant COX-2 enzyme as well as human blood monocytes as source of COX-2. As outlined earlier, acetaminophen’s potency as COX-2 inhibitor strongly depends on the oxidant/antioxidant status of the surrounding system (7, 23, 24). Thus, the presence of various enzymatic and nonenzymatic antioxidant components in human plasma (33) may explain why acetaminophen elicited the most pronounced COX-2 inhibition in human whole blood as compared to the two other experimental systems. Apart from these factors, evidence for COX-2 inhibition by acetaminophen is also supported by previous findings showing that acetaminophen like indomethacin and aspirin confers the synthesis of 18R-hydroxyeicosapentaenoic acid (HEPE) and 15RHEPE through interaction with the COX-2 enzyme (34). Acetaminophen’s only moderate inhibition of COX-1 in probands is reflected by its weak antiplatelet activity and good gastrointestinal safety. In fact, previous data suggest that only an excess of 95% inhibition of serum TxB2 significantly affects platelet function (35). In line with our data, acetaminophen at a single oral dose of 1000 mg (14) or up to 1950 mg (15) does not inhibit platelet function. Clinical trials reporting inhibition of platelet function by acetaminophen used high parenteral doses of the drug (36, 37). The same dose-dependent pattern applies for acetaminophen’s gastrointestinal safety. Accordingly, higher doses of acetaminophen confer higher rates of gastrointestinal events (e.g., dyspepsia) compared with lower doses probably due to a more pronounced COX-1 inhibition. In contrast to several short-term randomized trials, outcomes of epidemiological studies suggested that acetaminophen, at daily doses ⬎2 or ⬎2.6 g, increases the risk of severe upper gastrointestinal side effects, including gastrointestinal bleeding or perforation (14, 38). However, these conclusions cannot be inferred
The FASEB Journal
HINZ ET AL.
from this data because of several confounding factors, including the more common use of acetaminophen in patients at higher risk for gastrointestinal complications (39). Thus, acetaminophen is still regarded as free of major gastrointestinal toxicity (39 – 41). Nevertheless, large-scale randomized gastrointestinal outcome trials should be performed in patients receiving long-term acetaminophen. Referring to the gastrointestinal safety of acetaminophen, another fact appears noteworthy. Accordingly, in addition to its significant lower impact on COX-1 than that of NSAIDs, physicochemical factors (lack of acidity) may be responsible for acetaminophen’s favorable gastric tolerability as compared to acidic NSAIDs. Thus, apart from its ability to suppress PG synthesis, NSAID’s ulcerogenic action in the stomach is conferred by a topical irritant effect on the epithelium, which is due to the phenomenon of “ion trapping,” i.e., to accumulation of acidic NSAIDs in gastric epithelial cells (42– 44). In this context acidic NSAIDs have been suggested to produce mucosal injury by uncoupling oxidative mitochondrial phosphorylation in epithelial cells, resulting in diminished cellular ATP production, cellular toxicity, and increased mucosal permeability (43). A COXindependent mechanism for the gastrointestinal toxicity of NSAIDs is further supported by findings in COX-1 knockout mice, which still develop gastric erosions in response to oral administration of indomethacin (45). Finally, the issue of whether the demonstrated selective ex vivo COX-2 inhibition in adult males constitutes a general event remains to be determined. In case of gender, however, the previously shown 22% higher clearance of acetaminophen in males as compared to females (46) is not considered to be of clinical importance. In reference to children, the target plasma concentration of acetaminophen for antipyresis is 10 –20 g/ml (i.e., 66 –132 mol/L; 47), which compares favorably with the range of acetaminophen concentrations leading to profound COX-2 inhibition in human whole blood of adults. Although the ideal plasma concentration for analgesia remains undefined, it is supposed to lie within the same range (48). However, current product information recommends a rectal acetaminophen dose of 10 –15 mg/kg for children, which yields peak plasma concentrations less than the cited pharmacological concentration, thus raising the question whether this dose is adequate for effective postoperative pain therapy (48). In this context a previous study has shown that in children an initial rectal acetaminophen dose of 40 mg/kg followed by 20 mg/kg doses every 6 h is needed to achieve target antipyretic plasma concentrations (49). With respect to elderly patients, previous data indicate no age-related changes in the rate and extent of absorption and plasma clearance of acetaminophen (50). A more general issue constitutes an age-related increase in COX-2 expression that has been recently observed in human mononuclear cells, with a 70% increase in the older age group (51). The issue whether this change may affect the COX-2-inhibitory action of drugs is SELECTIVE COX-2 INHIBITION BY ACETAMINOPHEN
unanswered and remains to be determined. However, a significant impact seems unlikely given that a virtually complete COX-2 inhibition was observed in a comparable group of elderly probands (50 to 60 years) following an 8-day treatment with antirheumatic doses of diclofenac (20). Collectively, our data show a substantial and selective inhibition of COX-2 in healthy volunteers. This finding stands in clear contrast to previous concepts claiming a minor and clinically irrelevant degree of peripheral COX inhibition by acetaminophen. COX-1 inhibition was minor and never achieved a 95% inhibition needed for suppression of platelet function. Our data suggest a reinvestigation of acetaminophen in terms of peripheral analgesic and antiinflammatory mechanisms. Moreover, in view of recent concepts linking long-term suppression of COX-2 to cardiovascular side effects, recent warnings for patients receiving COX-2 inhibitors should also be considered for those taking high daily doses of acetaminophen for prolonged periods.
REFERENCES 1. 2. 3. 4. 5.
6.
7. 8.
9. 10. 11. 12.
13.
Schnitzer, T. J. (2002) Update of ACR guidelines for osteoarthritis: role of the coxibs. J. Pain Symptom Manage. 23 (Suppl. 4), S24 –S30 Botting, R. M. (2000) Mechanism of action of acetaminophen: is there a cyclooxygenase 3? Clin. Infect. Dis. 31 (Suppl. 5), S202–S210 Flower, R. J., and Vane, J. R. (1972) Inhibition of prostaglandin synthetase in brain explains the anti-pyretic activity of paracetamol (4-acetamidophenol). Nature 240, 410 – 411 Lanz, R., Polster, P., and Brune, K. (1986) Antipyretic analgesics inhibit prostaglandin release from astrocytes and macrophages similarly. Eur. J. Pharmacol. 130, 105–109 Swierkosz, T. A., Jordan, L., McBride, M., McGough, K., Devlin, J., and Botting, R. M. (2002) Actions of paracetamol on cyclooxygenases in tissue and cell homogenates of mouse and rabbit. Med. Sci. Monit. 8, BR496 –BR503 Warner, T. D., Vojnovic, I., Giuliano, F., Jimenez, R., BishopBailey, D., and Mitchell, J. A. (2004) Cyclooxygenases 1, 2, and 3 and the production of prostaglandin I2: investigating the activities of acetaminophen and cyclooxygenase-2-selective inhibitors in rat tissues. J. Pharmacol. Exp. Ther. 310, 642– 647 Graham, G. G., and Scott, K. F. (2005) Mechanism of action of paracetamol. Am. J. Ther. 12, 46 –55 Chandrasekharan, N. V., Dai, H., Roos, K. L., Evanson, N. K., Tomsik, J., Elton, T. S., and Simmons, D. L. (2002) COX-3, a cyclooxygenase-1 variant inhibited by acetaminophen and other analgesic/antipyretic drugs: cloning, structure, and expression. Proc. Natl. Acad. Sci. U. S. A. 99, 13926 –13931 Kis, B., Snipes, J. A., and Busija, D. W. (2005) Acetaminophen and the cyclooxygenase-3 puzzle: sorting out facts, fictions, and uncertainties. J. Pharmacol. Exp. Ther. 315, 1–7 Dinchuk, J. E., Liu, R. Q., and Trzaskos, J. M. (2003) COX-3: in the wrong frame in mind. Immunol. Lett. 86, 121 Schwab, J. M., Beiter, T., Linder, J. U., Laufer, S., Schulz, J. E., Meyermann, R., and Schluesener, H. J. (2003) COX-3—a virtual pain target in humans? FASEB J. 17, 2174 –2175 Qin, N., Zhang, S. P., Reitz, T. L., Mei, J. M., and Flores, C. M. (2005) Cloning, expression, and functional characterization of human cyclooxygenase-1 splicing variants: evidence for intron 1 retention. J. Pharmacol. Exp. Ther. 315, 1298 –1305 Garcia Rodriguez, L. A., and Hernandez-Diaz, S. (2001) Relative risk of upper gastrointestinal complications among users of acetaminophen and nonsteroidal anti-inflammatory drugs. Epidemiology 12, 570 –576
389
14. 15.
16. 17.
18.
19.
20.
21. 22.
23. 24.
25. 26.
27.
28.
29.
30.
31.
390
Mielke, C. H. Jr. (1981) Comparative effects of aspirin and acetaminophen on hemostasis. Arch. Intern. Med. 141, 305–310 Catella-Lawson, F., Reilly, M. P., Kapoor, S. C., Cucchiara, A. J., DeMarco, S., Tournier, B., Vyas, S. N., and FitzGerald, G. A. (2001) Cyclooxygenase inhibitors and the antiplatelet effects of aspirin. N. Engl. J. Med. 345, 1809 –1817 Jenkins, C., Costello, J., and Hodge, L. (2004) Systematic review of prevalence of aspirin induced asthma and its implications for clinical practice. BMJ 328, 434 Patrignani, P., Panara, M. R., Greco, A., Susco, O., Natoli, C., Iacobelli, S., Cipollone, F., Ganci, A., Creminon, C., Maclouf, J., and Patrono, C. (1994) Biochemical and pharmacological characterization of the cyclooxygenase activity of human blood prostaglandin endoperoxide synthases. J. Pharmacol. Exp. Ther. 271, 1705–1712 Hinz, B., Brune, K., and Pahl, A. (2000) Cyclooxygenase-2 expression in lipopolysaccharide-stimulated human monocytes is modulated by cyclic AMP, prostaglandin E2, and nonsteroidal anti-inflammatory drugs. Biochem. Biophys. Res. Commun. 278, 790 –796 Lau, G. S., and Critchley, J. A. (1994) The estimation of paracetamol and its major metabolites in both plasma and urine by a single high-performance liquid chromatography assay. J. Pharm. Biomed. Anal. 12, 1563–1572 Hinz, B., Dormann, H., and Brune, K. (2006) More pronounced inhibition of cyclooxygenase 2, increase in blood pressure, and reduction of heart rate by treatment with diclofenac compared with celecoxib and rofecoxib. Arthritis Rheum. 54, 282–291 Boardman, P. L., and Hart, F. D. (1967) Clinical measurement of the anti-inflammatory effects of salicylates in rheumatoid arthritis. BMJ 4, 264 –268 Ring, E. F., Collins, A. J., Bacon, P. A., and Cosh, J. A. (1974) Quantitation of thermography in arthritis using multi-isothermal analysis. II. Effect of nonsteroidal anti-inflammatory therapy on the thermographic index. Ann. Rheum. Dis. 33, 353–356 Ouellet, M., and Percival, M. D. (2001) Mechanism of acetaminophen inhibition of cyclooxygenase isoforms. Arch. Biochem. Biophys. 387, 273–280 Boutaud, O., Aronoff, D. M., Richardson, J. H., Marnett, L. J., and Oates, J. A. (2002) Determinants of the cellular specificity of acetaminophen as an inhibitor of prostaglandin H2 synthases. Proc. Natl. Acad. Sci. U. S. A. 99, 7130 –7135 Skjelbred, P., and Lokken, P. (1979) Paracetamol versus placebo: effects on post-operative course. Eur. J. Clin. Pharmacol. 15, 27–33 Bjornsson, G. A., Haanaes, H. R., and Skoglund, L. A. (2003) A randomized, double-blind crossover trial of paracetamol 1000 mg four times daily vs ibuprofen 600 mg: effect on swelling and other postoperative events after third molar surgery. Br. J. Clin. Pharmacol. 55, 405– 412 Ferreira, S. H., Lorenzetti, B. B., and Correa, F. M. (1978) Blockade of central and peripheral generation of prostaglandins explains the antialgic effect of aspirin like drugs. Pol. J. Pharmacol. Pharm. 30, 133–140 Honore, P., Buritova, J., and Besson, J. M. (1995) Aspirin and acetaminophen reduced both Fos expression in rat lumbar spinal cord and inflammatory signs produced by carrageenin inflammation. Pain 63, 365–375 Smith, C. J., Zhang, Y., Koboldt, C. M., Muhammad, J., Zweifel, B. S., Shaffer, A., Talley, J. J., Masferrer, J. L., Seibert, K., and Isakson, P. C. (1998) Pharmacological analysis of cyclooxygenase-1 in inflammation. Proc. Natl. Acad. Sci. U. S. A. 95, 13313– 13318 Cannon, C. P., Curtis, S. P., FitzGerald, G. A., Krum, H., Kaur, A., Bolognese, A., Reicin, A. S., Bombardier, C., Weinblatt, M. E., van der Heijde, D., Erdmann, E., and Laine, L., for the MEDAL Steering Commitee (2006) Cardiovascular outcomes with etoricoxib and diclofenac in patients with osteoarthritis and rheumatoid arthritis in the multinational etoricoxib and diclofenac arthritis long-term (MEDAL) programme: a randomised comparison. Lancet 368, 1771–1781 Forman, J. P., Stampfer, M. J., and Curhan, G. C. (2005) Non-narcotic analgesic dose and risk of incident hypertension in US women. Hypertension 46, 500 –507
Vol. 22
February 2008
32.
33. 34.
35. 36. 37.
38.
39. 40. 41. 42.
43.
44. 45.
46. 47. 48.
49.
50. 51.
Chan, A. T., Manson, J. E., Albert, C. M., Chae, C. U., Rexrode, K. M., Curhan, G. C., Rimm, E. B., Willett, W. C., and Fuchs, C. S. (2006) Nonsteroidal antiinflammatory drugs, acetaminophen, and the risk of cardiovascular events. Circulation 113, 1578 –1587 Halliwell, B., and Gutteridge, J. M. (1990) The antioxidants of human extracellular fluids. Arch. Biochem. Biophys. 280, 1– 8 Serhan, C. N., Clish, C. B., Brannon, J., Colgan, S. P., Chiang, N., and Gronert, K. (2000) Novel functional sets of lipid-derived mediators with antiinflammatory actions generated from omega-3 fatty acids via cyclooxygenase 2-nonsteroidal antiinflammatory drugs and transcellular processing. J. Exp. Med. 192 (Suppl. 8), 1197–1204 Reilly, I. A., and FitzGerald, G. A. (1987) Inhibition of thromboxane formation in vivo and ex vivo: implications for therapy with platelet inhibitory drugs. Blood 69, 180 –186 Nieme, T. T., Backman, J. T., Syrja¨la¨, M. T., Vinikka, L. U., and Rosenberg, P. H. (2000) Platelet dysfunction after intravenous ketorolac or propacetamol. Acta Anaesthesiol. Scand. 44, 69 –74 Munsterhjelm, E., Munsterhjelm, N. M., Niemi, T. T., Ylikorkala, O., Neuvonen, P. J., and Rosenberg, P. H. (2005) Dose-dependent inhibition of platelet function by acetaminophen in healthy volunteers. Anesthesiology 103, 712–717 Rahme, E., Pettitt, D., and LeLorier, J. (2002) Determinants and sequelae associated with utilization of acetaminophen versus traditional nonsteroidal antiinflammatory drugs in an elderly population. Arthritis Rheum. 46, 3046 –3054 Abramson, S. B. (2002) Et tu, acetaminophen? Arthritis Rheum. 46, 2831–2835 Bannwarth, B. (2004) Gastrointestinal safety of paracetamol: is there any cause for concern? Expert Opin. Drug Saf. 3, 269 –272 Graham, G. G., Scott, K. F., and Day, R. O. (2005) Tolerability of paracetamol. Drug Saf. 28, 227–240 McCormack, K., and Brune, K. (1987) Classical absorption therapy and the development of gastric mucosal damage associated with the non-steroidal anti-inflammatory drugs. Arch. Toxicol. 60, 261–269 Somasundaram, S., Rafi, S., Hayllar, J., Sigthorsson, G., Jacob, M., Price, A. B., Macpherson, A., Mahmod, T., Scott, D., Wrigglesworth, J. M., and Bjarnason, I. (1997) Mitochondrial damage: a possible mechanism of the “topical” phase of NSAID induced injury to the rat intestine. Gut 41, 344 –353 Wallace, J. L. (1997) Nonsteroidal anti-inflammatory drugs and gastroenteropathy: the second hundred years. Gastroenterology 112, 1000 –1016 Langenbach, R., Morham, S. G., Tiano, H. F., Loftin, C. D., Ghanayem, B. I., Chulada, P. C., Mahler, J. F., Lee, C. A., Goulding, E. H., Kluckman, K. D., Kim, H. S., and Smithies, O. (1995) Prostaglandin synthase 1 gene disruption in mice reduces arachidonic acid-induced inflammation and indomethacin-induced gastric ulceration. Cell 83, 483– 492 Miners, J. O., Attwood, J., and Birkett, D. J. (1983) Influence of sex and oral contraceptive steroids on paracetamol metabolism. Br. J. Clin. Pharmacol. 16, 503–509 Rumack, B. H. (1978) Aspirin vs. acetaminophen: a comparative view. Pediatrics 62, 943–946 Mantzke, U. S., and Brambrink, A. M. (2002) Paracetamol in childhood. Current state of knowledge and indications for a rational approach to postoperative analgesia. Anaesthesist 51 (Suppl. 9), 735–746 Birmingham, P. K., Tobin, M. J., Fisher, D. M., Henthorn, T. K., Hall, S. C., and Cote´, C. J. (2001) Initial and subsequent dosing of rectal acetaminophen in children: a 24-hour pharmacokinetic study of new dose recommendations. Anesthesiology 94, 385–389 Triggs, E. J., and Nation, R. L. (1975) Pharmacokinetics in the aged: a review. J. Pharmacokinet. Biopharm. 3, 387– 418 Kang, K. B., Van Der Zypp, A., Iannazzo, L., and Majewski, H. (2006) Age-related changes in monocyte and platelet cyclooxygenase expression in healthy male humans and rats. Transl. Res. 148 (Suppl. 6), 289 –294
The FASEB Journal
Received for publication March 27, 2007. Accepted for publication August 6, 2007.
HINZ ET AL.
