Professor G. Geisslinger, Pharmazentrum Frankfurt, Institut für Klinische Pharmakologie, Klinikum der Johann-Wolfgang-Goethe-Universität Frankfurt, Theodor-Stern-Kai 7, 60590 Frankfurt/Main, Germany. E-mail: email@example.com
Cyclo-oxygenase-2-selective inhibitors produce less gastric damage than conventional non-steroidal anti-inflammatory drugs. Valdecoxib is a new orally administered cyclo-oxygenase-2-selective inhibitor, recently approved for use in osteoarthritis, rheumatoid arthritis and primary dysmenorrhoea in the USA. The drug has been evaluated in more than 60 clinical studies involving more than 14 000 patients and healthy volunteers. The analgesic efficacy of valdecoxib at a dose of 10 mg once daily in both osteoarthritis and rheumatoid arthritis is superior to that of placebo and similar to that of traditional non-steroidal anti-inflammatory drugs. Valdecoxib is effective in single doses of up to 40 mg for the alleviation of acute menstrual pain and has a rapid onset of action (within 30 min) and a long duration of analgesia (up to 24 h). Valdecoxib is well tolerated and has safety advantages compared with traditional non-steroidal anti-inflammatory drugs in terms of less gastrointestinal toxicity and a lack of an effect on platelet function. The incidence of adverse effects involving the kidney (fluid retention, oedema and hypertension) is similar to that of non-selective, non-steroidal anti-inflammatory drugs.
Conventional non-steroidal anti-inflammatory drugs (NSAIDs) are the mainstay of treatment of arthritis and pain, but their long-term use can lead to severe and occasionally life-threatening adverse effects involving the gastrointestinal tract, which limit their therapeutic benefit.1,2 Progress towards safer analgesic and anti-inflammatory therapy has been aided by the identification of two isoforms of the enzyme cyclo-oxygenase (COX): COX-1 and COX-2.3–5 Although COX-1 is constitutively expressed in most tissues (e.g. in the gastric mucosa and platelets), COX-2 appears to be highly inducible at sites of inflammation and in cancer. As they are non-selective inhibitors of COX, traditional NSAIDs can limit inflammation, but also interfere with some COX-1-mediated physiological mechanisms, resulting in gastrointestinal damage and bleeding-related disorders.6,7
The differences between COX-1 and COX-2 in terms of expression and function have led to the research and development of drugs able to selectively inhibit COX-2. The aim of this approach is to identify agents with the required antipyretic, analgesic and anti-inflammatory efficacy shown by NSAIDs, but with a COX-1-sparing effect.7 Following the Food and Drug Administration's approval of the first of these COX-2-selective inhibitors, namely celecoxib and rofecoxib, several other ‘coxibs’ have reached or will shortly reach the market. One of the newer agents is valdecoxib [4-(5-methyl-3-phenyl-4-isoxazolyl)benzenesulphonamide], which has recently been approved in the USA for the treatment of the signs and symptoms of osteoarthritis, rheumatoid arthritis and primary dysmenorrhoea.
Valdecoxib has been evaluated in more than 60 clinical studies involving more than 14 000 patients and healthy volunteers. This article summarizes published and yet unpublished (on file) pharmacodynamic and pharmacokinetic data on valdecoxib, as well as the clinical experience gained with the drug in the treatment of patients with osteoarthritis, rheumatoid arthritis, primary dysmenorrhoea and acute pain.
Mechanism of action of valdecoxib
Valdecoxib, a diaryl-substituted derivative of isoxazole, binds non-covalently to COX-2, forming a tight and relatively stable enzyme–inhibitor complex. In vitro assays have confirmed that valdecoxib is a potent inhibitor of prostaglandin E2 production by human recombinant COX-2 [concentration required for 50% inhibition of enzyme activity (IC50) = 0.005 µm] when compared with its effect on COX-1 (IC50 = 140 µm). The selectivity for COX-2 has also been shown ex vivo in human whole-blood assays, an assay procedure providing a valid index of selectivity because of the binding of the drug to plasma proteins and the presence of endogenous arachidonic acid formation. However, such data are highly variable and need to be interpreted with caution because of inter-assay and inter-laboratory differences. For valdecoxib and celecoxib, the COX-1/COX-2 ratios determined in a single whole-blood assay in the same laboratory were 29 and 40, respectively.8 These values are at least 10-fold greater than the COX-1/COX-2 ratios obtained with non-selective NSAIDs.9
The primary hydroxylated metabolite of valdecoxib, SC-66905, which accounts for 10–15% of the parent compound in human plasma (Pharmacia, on file 2000), has lower COX-2 inhibitory activity than that of valdecoxib and does not contribute significantly to the clinical effects of valdecoxib. Valdecoxib and its metabolite SC-66905 are also the active moieties of a new parenteral COX-2-selective inhibitor, parecoxib sodium.10
Valdecoxib is effective in several rat in vivo models of inflammation and nociception, whilst showing a good gastrointestinal safety profile (Table 1). In particular, in the rat carrageenan air pouch model, the levels of COX-2-mediated prostaglandin E2 in the air pouch exudates were significantly reduced by valdecoxib, while the production of prostaglandin E2 in the stomach mucosa mediated by COX-1 showed no significant change.11 In the carrageenan paw oedema model, valdecoxib reduced inflammatory paw oedema by up to 70% compared with baseline values, an effect comparable with that of naproxen (67%).12 Furthermore, in the Hargreaves hyperalgesia model, valdecoxib was able to reverse hyperalgesia by 97%, compared with 74% for naproxen.13 In the rat adjuvant-induced arthritis model of chronic inflammation, the anti-inflammatory effect of valdecoxib was similar to that of indometacin, with maximal reductions in the swelling of the hind limb of 95% for the former and 93% for the latter.14 Valdecoxib, at doses of 3, 10, 30 and 100 mg/kg, was also effective in reversing hyperalgesia and allodynia in a Brennan model of post-operative pain, in which nociception was produced by surgical incisions.15 A maximal reduction in pain response was achieved at the lowest dose tested (3 mg/kg), indicating that valdecoxib has a marked anti-nociceptive effect in acute tissue injury.
Table 1. Anti-inflammatory and anti-nociceptive effects of valdecoxib and celecoxib in various rat models in vivo8,11,60
The efficacy of cyclo-oxygenase (COX) inhibitors is expressed as ED50 (the dose of drug required to elicit 50% of the maximal response).
Air pouch prostaglandins reflect mainly COX-2 activity in vivo.
Gastric prostaglandins reflect mainly COX-1 activity in vivo.
Pharmacokinetics of valdecoxib in the human
Young healthy subjects
Valdecoxib has good oral bio-availability (about 83%) and a minimal first-pass effect. In fasting subjects, valdecoxib administered orally is rapidly absorbed and achieves maximal plasma concentrations in approximately 3 h (Tmax). Although a high-fat diet delays the absorption of valdecoxib by 1–2 h, there are no significant effects of a meal on the peak plasma concentration (Cmax) or the area under the concentration–time curve (AUC).16
The concomitant administration of valdecoxib with antacids (aluminium–magnesium hydroxide) has no significant effect on the rate and extent of absorption of the drug. Although data on the pharmacokinetics of valdecoxib during more profound inhibition of gastric acid secretion are not available, one would not expect a change in valdecoxib absorption when given concomitantly with H2-receptor antagonists or proton pump inhibitors. This conclusion is supported by studies showing no significant changes in the dissolution rate and octanol/water partition coefficient of valdecoxib over a wide pH range (pH 2–9), and the good long-term stability (over 12 weeks) of valdecoxib solutions in this pH range (Pharmacia, on file 2001).
The Cmax and AUC values for valdecoxib show dose linearity with single doses of up to 400 mg.17 On multiple dosing, AUC increases in a non-linear manner with doses above 20 mg/day. At higher doses, there is a 25–45% disproportionate increase in AUC with dose, which is not clinically significant. Steady-state plasma concentrations of valdecoxib are achieved on the fourth day after commencing treatment.
The total plasma clearance of valdecoxib, measured after a 10-mg single intravenous infusion, is approximately 6.0 L/h, with a steady-state volume of distribution of approximately 54.5 L. Like many traditional NSAIDs, valdecoxib is highly protein bound (> 98%), and the percentage bound is independent of the total plasma concentration. Valdecoxib selectively partitions into erythrocytes, with an average erythrocyte/plasma ratio of about 4 : 1, which is independent of time and plasma concentration.16
Valdecoxib is eliminated predominantly via hepatic metabolism, with less than 5% of the dose excreted unchanged in the urine and faeces. About 70% of the dose is excreted in the urine as metabolites. Valdecoxib is metabolized via the cytochrome P450 enzymes, as well as via non-cytochrome-dependent reactions. Within the cytochrome P450 system, the P450 3A4 isozyme is primarily involved. However, P450 2C9 also plays a role. Hydroxylation of the methyl group by CYP3A4 produces a pharmacologically active (as a COX-2 inhibitor) metabolite SC-66905. This metabolite undergoes further metabolism via aryl hydroxylation, leading to the formation of a carboxylic acid derivative (Figure 1). Glucuronidation of valdecoxib and its metabolites accounts for 20% of the drug products excreted. The elimination half-life of valdecoxib (t1/2) is approximately 8–11 h. The role of genetic polymorphisms in drug-metabolizing enzymes on the bio-availability of valdecoxib is not known so far.
The pharmacokinetic properties of valdecoxib are summarized in Table 2, and compared with those of two other COX-2-selective inhibitors, celecoxib and rofecoxib.
Table 2. Pharmacokinetics and clinical applications of valdecoxib in comparison with celecoxib and rofecoxib
AUC, area under the concentration–time curve; CYP, cytochrome P450; GFR, glomerular filtration rate.
* Absolute bio-availability studies in humans have not been conducted; the values were obtained from pharmacokinetic studies in dogs.61
Time to reach maximum plasma concentration (Tmax, h)
Elimination half-life (t1/2, h)
Volume of distribution (L)
Extent of binding to plasma proteins (%)
Main metabolic pathway in liver metabolism
CYP3A4, CYP2C9 and P450-independent pathways
Influence of hepatic impairment (Child–Pugh class A and B)
AUC 19–130% higher
AUC 26–146% higher
AUC 30–69% higher
Influence of renal insufficiency (GFR, 35–60 mL/min)
AUC 43% lower
Recommended daily doses (mg)
In rheumatoid arthritis
In acute pain and primary dysmenorrhoea
Up to 40
Up to 600 on day 1, then 400 afterwards
Up to 50
Age-related changes in hepatic and renal function, as well as in the volume of distribution, lead to higher plasma concentrations of valdecoxib in the elderly. Studies with valdecoxib (10 mg b.d.) in healthy elderly subjects showed a significant reduction in plasma clearance and a 35–41% increase in AUC0−12 h per kilogram of body weight.16 The weight-adjusted steady-state AUC is 16% higher in elderly females than in elderly males (Pharmacia, on file 2000). These data suggest that therapy with valdecoxib in elderly patients, especially in those with a low body weight (< 50 kg), should be initiated with the lowest recommended dose.
Patients with hepatic impairment
Clinical experience using valdecoxib in patients with hepatic dysfunction is limited. However, data have been reported on the effect of liver dysfunction on the pharmacokinetics of the pro-drug parecoxib sodium, which is rapidly and almost completely converted to valdecoxib via enzymatic hydrolysis in the liver.18
In patients with mild hepatic impairment (Child–Pugh class A) undergoing long-term treatment with parecoxib sodium (10 mg b.d.), Cmax and AUC0−12 h of valdecoxib were increased by 26% and 19%, respectively, compared with healthy control subjects. On the other hand, in patients with moderate liver dysfunction (Child–Pugh class B), the same therapy resulted in a 142% increase in Cmax and a 130% increase in AUC0−12 h (Pharmacia, on file 1999). As the rate of hydrolysis of parecoxib to valdecoxib remained unchanged, the observed increase in valdecoxib concentration was attributed to reduced hepatic elimination. Therefore, as in the case of other ‘coxibs’, valdecoxib treatment in patients with moderate hepatic impairment should begin with small doses and patients should be closely monitored. In mild hepatic impairment, there does not appear to be a need for dose adjustment. As no data for patients with severe hepatic dysfunction (Child–Pugh class C) are available, the use of valdecoxib in these patients is not recommended.
Patients with renal impairment
The pharmacokinetics of valdecoxib have been studied in patients with varying degrees of renal impairment. Following a single oral dose of 20 mg valdecoxib in patients with severe renal dysfunction (creatinine clearance ≤ 16.7 mL/min/1.73 m2) and those undergoing haemodialysis with end-stage renal disease, there was only a slight reduction (23%) in the mean plasma clearance compared with healthy control subjects. The average ratio of valdecoxib AUC0–inf to SC-66905 AUC0–inf showed no significant change, suggesting that conversion of valdecoxib to SC-66905 is not altered in renal impairment (Pharmacia, on file 2000). These findings confirm that valdecoxib and its active metabolite SC-66905 are eliminated predominantly via hepatic metabolism, and that the renal route plays a minor role in the excretion of these compounds. In patients undergoing haemodialysis (n = 8), only 2% of the administered dose of valdecoxib appeared in the dialysate, probably because of the high protein binding and low renal clearance of valdecoxib. Although the available pharmacokinetic data indicate that no significant adjustment in the dosage is required in patients with renal insufficiency, the use of valdecoxib in advanced renal disease is not recommended, as selective and non-selective NSAIDs have been associated with worsening renal function (see below).
As valdecoxib is eliminated predominantly via hepatic metabolism, involving cytochrome P450 3A4 and 2C9, the prototype inhibitors and substrates of these isozymes have been extensively studied in various drug interaction trials with valdecoxib and its pro-drug parecoxib sodium.
In healthy volunteers, pre-treatment with valdecoxib (40 mg b.d. for 7 days) showed no interaction with a single oral dose of the CYP3A4 substrate midazolam (10 mg). Similarly, administration of the pro-drug parecoxib sodium (single intravenous dose of 40 mg) showed no pharmacokinetic or pharmacodynamic interactions with midazolam (0.07 mg/kg intravenously)19 or the CYP2C9 substrate propofol (2 mg/kg intravenously).20 No pharmacokinetic or pharmacodynamic (pupil diameter and respiratory rate) alterations were seen when the CYP3A4 substrates alfentanil (15 µg/kg) and fentanyl (5 µg/kg) were given to subjects receiving 40 mg parecoxib sodium as a single intravenous dose.21 However, in subjects pre-treated with the CYP3A4 inhibitor ketoconazole (200 mg b.d. for 7 days) and the CYP2C9/3A4 inhibitor fluconazole (200 mg/day for 7 days), the plasma concentrations of valdecoxib, following single-dose administration (20 mg), increased significantly by 38% and 62%, respectively.16
A drug interaction study with valdecoxib (40 mg b.d. for 7 days) and the CYP2C9/3A4 substrate warfarin (1–8 mg/day) demonstrated a statistically significant increase in plasma concentrations of R-warfarin and S-warfarin (15% and 12%, respectively) and a small, but statistically significant, increase in the prothrombin time expressed as the International Normalized Ratio.16 This finding suggests that anticoagulant therapy with warfarin in patients receiving valdecoxib should be closely monitored, particularly during the first few weeks after beginning valdecoxib treatment or when changing the dose of valdecoxib.
The interaction potential of valdecoxib (10 mg b.d. for 7 days) with regard to CYP3A4 has also been tested in patients with non-insulin-dependent diabetes mellitus receiving the anti-diabetic agent glyburide (5 mg once daily or 10 mg twice daily). Although no significant changes in the pharmacokinetics of glyburide were observed, there was a slight, non-significant, decrease in average blood glucose concentrations (4% for 5 mg/day and 9% for 10 mg b.d.) in the group receiving valdecoxib when compared with that receiving placebo. The reduction in blood glucose, however, was not clinically significant (Pharmacia, on file 1999).
A drug interaction study with valdecoxib (40 mg b.d. for 7 days) and dextromethorphan (30 mg single dose), which is metabolized primarily by CYP2D6 and, to a lesser extent, by CYP3A4, showed an increase in dextromethorphan AUC by 228% and Cmax by 111%. This effect, however, was less significant than that observed with the CYP2D6 inhibitor paroxetine (20 mg b.d. for 7 days) used as a positive control (increase in AUC by 2264% and Cmax by 601%), suggesting that valdecoxib is a weak inhibitor of CYP2D6. In these investigations, the inhibition or induction of the CYP3A4-mediated metabolism of dextromethorphan to 3-methoxymorphinan was not seen (Pharmacia, on file 2000).
As valdecoxib may be associated with impaired renal function (see below), drug interaction studies have been carried out with drugs undergoing marked renal excretion. The concomitant administration of valdecoxib (40 mg b.d. for 6.5 days) with lithium carbonate (450 mg controlled-release tablets b.d. for 6.5 days), in particular, led to a reduced renal clearance of lithium (30%), with a 34% increase in its serum exposure (AUC), but had no effect on the plasma concentration of valdecoxib.16 Therefore, when lithium carbonate is taken together with valdecoxib, the lithium concentrations should be closely monitored, especially at the commencement of valdecoxib treatment and when changing the dose of valdecoxib.
Clinical data on the interaction between valdecoxib at various doses and methotrexate in patients with rheumatoid arthritis have recently been published.22 The concomitant administration of an intramuscular dose of methotrexate (5–20 mg once weekly) during treatment with valdecoxib (40 mg b.d. for 6.5 days) had no significant effect on the plasma AUC of methotrexate when compared with placebo. A 19% reduction in the renal clearance of methotrexate with valdecoxib (40 mg twice daily) was compensated by a small increase in the non-renal plasma clearance of methotrexate, resulting in little or no change in its AUC. Low doses of valdecoxib (10 mg b.d. for 6 days) had no effect on the pharmacokinetics of orally administered methotrexate.
Valdecoxib efficacy in the management of acute pain
The analgesic efficacy, duration of action and safety of a single oral dose of valdecoxib have been evaluated in a randomized, single-centre, double-blind, placebo-controlled trial in 284 adults following extraction of two ipsilateral impacted third molars and bony resection.23 Valdecoxib (10, 20, 40 or 80 mg) or placebo was administered 60–75 min prior to surgery. The surgical procedure with local anaesthesia lasted for 20–30 min. Efficacy was assessed using pain intensity levels, the percentage of patients requiring rescue medication over 24 h and the patient's global evaluation. In general, the duration and magnitude of analgesia for all pre-operatively administered doses of valdecoxib were significantly greater than for placebo (P < 0.001). Valdecoxib was also significantly superior to placebo in terms of the percentage of patients requiring rescue medication (67% for 10 mg vs. 57% for 20 mg vs. 32% for 40 mg vs. 41% for 80 mg valdecoxib vs. 95% for placebo), and the percentage of patients reporting good or excellent relief from pain (78–93% for 10–80 mg valdecoxib vs. 48% for placebo). A dose-dependent analgesic effect was observed up to a dose of 40 mg.
Two studies have been conducted comparing the analgesic efficacy and safety of valdecoxib (20 or 40 mg) with oxycodone 10 mg/acetaminophen 1000 mg in patients following oral surgery. Valdecoxib (40 mg) showed a rapid onset of analgesia (34 min) in both studies, and the magnitude of pain relief was comparable with the oxycodone/acetaminophen combination, but exhibited a significantly longer duration of analgesic action. Pooled safety data demonstrated that valdecoxib had a tolerability profile superior to that of the oxycodone/acetaminophen combination and similar to that of placebo.24
The efficacy of a single dose of 40 mg valdecoxib has been compared with 50 mg rofecoxib and placebo in a randomized, double-blind, placebo-controlled clinical trial in 203 patients with moderate to severe pain following the extraction of two or more third molars with bone removal.25 After surgery, patients received a single dose of 40 mg valdecoxib (n = 80), 50 mg rofecoxib (n = 82) or placebo (n = 41). Analgesia was assessed over 24 h following surgery, and rescue medication was available on request. When compared with rofecoxib, patients receiving valdecoxib experienced a shorter time to perceptible pain relief (34 min vs. 55 min, P ≤ 0.05), a faster onset of analgesia (43 min vs. 60 min) and a markedly improved pain relief score (P < 0.05) in the period 0.75–24 h post-dose. Moreover, significantly more patients (68%) who received a single dose of 40 mg valdecoxib rated their study medication as ‘good or excellent’, compared with only 44% in the rofecoxib 50 mg group (P ≤ 0.01).
Several clinical studies, involving patients undergoing hip arthroplasty26 and bunionectomy,27 have shown that valdecoxib is an effective opioid-sparing analgesic in the management of post-operative pain.
Patients undergoing hip arthroplasty were randomized to receive valdecoxib (20 or 40 mg b.d.) or placebo every 12 h starting 1–3 h before surgery. Over a 48-h period after surgery, patients receiving 20 or 40 mg valdecoxib required less morphine (43% and 41%, respectively) than those receiving placebo (P < 0.001). Pain intensity levels and patient satisfaction with the medication were also significantly improved with both doses of valdecoxib compared with placebo.
In a bunionectomy study, a single dose of valdecoxib (20, 40 or 80 mg) or placebo administered 1 h (average, 45–75 min) before surgery significantly delayed the onset of post-operative pain (8 h for valdecoxib vs. 3.5 h for placebo).27 Significantly more patients receiving valdecoxib than those receiving placebo (12–29% with 20–80 mg valdecoxib vs. 0.00% with placebo) required no additional analgesics during the 24-h period after surgery.
The results of a large, multi-centre, double-blind, placebo-controlled trial of valdecoxib (n = 209) in post-operative pain management following knee surgery have been presented recently at the meeting of the American Academy of Orthopedic Surgeons (February 2002, Dallas, TX, USA). In contrast with the studies mentioned above, treatment with valdecoxib (20 or 40 mg) or placebo in this trial was initiated post-operatively, as soon as the patient could take oral medication, with follow-up doses 12, 24 and 36 h after the first dose or until discontinuation of morphine. Efficacy assessment was based on the dose of morphine administered by pump during patient-controlled analgesia in the first 24–48-h period after the first dose of valdecoxib. Patients receiving valdecoxib consumed smaller amounts of morphine than those receiving placebo. In particular, the reductions in morphine consumption with valdecoxib at doses of 20 mg and 40 mg were 16.3% and 24.2%, respectively. A higher percentage of patients treated with valdecoxib, 20 and 40 mg, in combination with morphine (79% and 84%, respectively) described their study medication as ‘good’ or ‘excellent’, compared with 70% of patients receiving morphine plus placebo.
Valdecoxib efficacy in osteoarthritis
The efficacy of valdecoxib in osteoarthritis of the hip or knee has been evaluated in two double-blind, randomized controlled trials in which a total of 1485 patients were treated for 3 months.28, 29 Multiple-dose regimens (5–20 mg/day) were administered for 12 weeks and compared with naproxen (500 mg b.d.). Patient evaluation was performed at baseline and after 2, 6 and 12 weeks of treatment, or on early withdrawal. Standard efficacy assessments included, but were not limited to, the Patient's and Physician's Global Assessment of Arthritis and WOMAC (Western Ontario and McMaster Universities), which are validated tools for the evaluation of osteoarthritis based on the separate assessment of pain, joint stiffness and physical function, as well as on the composite index derived from these parameters. In general, valdecoxib was superior to placebo in improving the clinical signs and symptoms of osteoarthritis. In terms of efficacy, valdecoxib, 10 mg once daily, was similar to naproxen at all assessment points, and superior to valdecoxib, 5 mg once daily. As no additional benefit was seen with valdecoxib, 20 mg once daily, 10 mg/day is the currently recommended dose for the treatment of osteoarthritis.
Valdecoxib efficacy in adult rheumatoid arthritis
The efficacy of valdecoxib in patients with rheumatoid arthritis has been evaluated according to a study design similar to that in the osteoarthritis trials mentioned above. In a multi-centre study, once-daily doses of 10, 20 and 40 mg of valdecoxib were evaluated vs. naproxen (500 mg b.d.) over a 12-week period in a total of 1089 patients.30 Efficacy assessment using the standard criteria of the American College of Rheumatology (ACR) responder index, the ACR-20, was made at baseline and after 2, 6 and 12 weeks of treatment, or on early termination. The percentage of patients improving according to the ACR-20 after 12 weeks of treatment was significantly greater in all groups receiving valdecoxib (46–49% for 10–40 mg) compared to placebo (32%, P < 0.001). Furthermore, the efficacy of valdecoxib, at all three dosages, was comparable with that of naproxen (500 mg b.d.) (44%). There was no additional benefit seen when the daily dose of valdecoxib was increased from 10 mg to 20 or 40 mg. This finding provides the basis for the recommended dose of 10 mg/day of valdecoxib in rheumatoid arthritis patients.
Valdecoxib efficacy in primary dysmenorrhoea
A randomized, double-blind, placebo-controlled, cross-over study (n = 120) has compared the analgesic efficacy of valdecoxib (20 and 40 mg b.d. according to need) with that of placebo and naproxen sodium (550 mg b.d. according to need) for the treatment of pain associated with primary dysmenorrhoea.31,32 The criterion for enrolment in the study was that patients should have a history of primary dysmenorrhoea with continual menstrual cramping pain of moderate or severe intensity during at least four of the previous 6 months. The treatment was initiated during the menses, at a time when pain due to menstrual cramping was unendurable. Additional doses every 8–12 h up to 72 h were allowed according to need. The main index of analgesic efficacy used was based on the sum of the pain intensity differences and the total pain relief score, which are standard measures of pain in clinical studies. The analgesic effect of valdecoxib (20 and 40 mg) was detectable 30–60 min after drug intake and was significantly superior to that of placebo for the first 8 and 12 h after the initial dose (20 mg, P < 0.01; 40 mg, P < 0.001). Valdecoxib, 20 mg and 40 mg, were comparable with naproxen sodium (550 mg) in both the magnitude and duration of action. Because at least 80% of the enrolled patients required only a single dose of valdecoxib (20 and 40 mg) during the first 24 h for the alleviation of menstrual pain, the recommended dose of valdecoxib in the treatment of primary dysmenorrhoea is 20 mg twice daily as needed.
Safety and gastrointestinal tolerability of valdecoxib
The safety and gastrointestinal tolerability of valdecoxib have been evaluated in all clinical trials carried out with this agent. In the oral surgery study, the safety data obtained after single oral doses of 10, 20, 40 and 80 mg valdecoxib (n = 227) showed no significant differences in the total incidence of adverse events when compared with the placebo-treated group (n = 57).23 The adverse reactions observed were typical of those associated with the oral surgical procedure and included headache, nausea, vomiting, somnolence, dizziness, facial oedema and occasionally earache.
Data on the long-term administration of valdecoxib at doses of 5–40 mg, pooled from three multi-centre studies with a total of 1480 patients with osteoarthritis of the hip and knee and with rheumatoid arthritis, confirm that the drug is safe when used in chronic treatment.28–30 The most common adverse effects, for which the incidence was at least 5% during the 12-week treatment period, were gastrointestinal symptoms (e.g. abdominal pain, diarrhoea, dyspepsia and nausea), headache and infection of the upper respiratory tract. The incidence of headache and upper respiratory tract infections at all doses of valdecoxib was similar to that for placebo, as was the incidence of gastrointestinal symptoms at valdecoxib doses of 5 and 10 mg. At higher doses (20–40 mg), gastrointestinal symptoms occurred more frequently with valdecoxib than with placebo, and the difference in the incidence of dyspepsia with 40 mg valdecoxib was statistically significant vs. placebo in rheumatoid arthritis patients (9.5% vs. 4.1%; P < 0.05). However, the incidence of gastrointestinal symptoms in all valdecoxib groups was significantly lower than that in the group receiving the traditional NSAID naproxen (n = 547).
In an open-label, multi-centre, long-term safety study of valdecoxib (40 mg b.d.), 21 osteoarthritis patients (10.0%) and 19 rheumatoid arthritis patients (10.2%) withdrew before study completion because of adverse events. The incidence of each observed event was less than one per 100 patient-years, with the exception of abdominal pain, where the occurrence rate was 1.7 (Pharmacia, on file 2000).
A more thorough evaluation of the gastrointestinal effects of valdecoxib has been carried out in endoscopic studies of up to 26 weeks' duration. In these studies, the incidence of gastrointestinal ulcers during treatment with valdecoxib (5, 10 and 20 mg daily) was compared with that with ibuprofen (800 mg t.d.s.), diclofenac sodium (75 mg b.d.) and naproxen (500 mg b.d.).33–36 Endoscopy of the upper gastrointestinal tract was performed before and after the first dose of study medication, as well as at the termination of the study (e.g. after 12 or 26 weeks of treatment). Patients who were not ulcer free at baseline endoscopy were not included in the trials. The primary end-point was the incidence of gastroduodenal, gastric and duodenal ulcers observed during endoscopy of the upper gastrointestinal tract.
The percentage of osteoarthritis patients with gastroduodenal ulcers after 12 weeks of treatment with valdecoxib (10 or 20 mg) was significantly lower (4.9% for 10 mg and 4.4% for 20 mg) than in the diclofenac group (17.2%) and the ibuprofen group (16.1%).36 The incidence of gastroduodenal ulcer with placebo was 6.5%.
In another trial in osteoarthritis patients (n = 1019), the incidence of gastroduodenal ulcers resulting from valdecoxib usage after 12 weeks (3%, 3% and 5% for 5 mg, 10 mg and 20 mg valdecoxib, respectively) was comparable with that of placebo (4%) and significantly lower than that in the naproxen group (10%).29
In a recently completed endoscopic evaluation of the gastrointestinal safety of supra-therapeutic doses of valdecoxib (20 mg and 40 mg b.d.) and naproxen (500 mg b.d.) over 14 weeks, carried out in osteoarthritis and rheumatoid arthritis patients, the incidence of ulcers in valdecoxib-treated patients was significantly lower (4% and 8%) than in the naproxen group (18%), and comparable with that previously observed with lower doses of valdecoxib.33
In general, these findings show that valdecoxib produces fewer ulcers than produced by non-selective NSAIDs. They are in agreement with safety data obtained with other COX-2-selective inhibitors, including celecoxib37,38 and rofecoxib,39–41 and are supported by experimental data in the rat showing that the inhibition of both COX-1 and COX-2 is required for NSAID-induced gastric injury.42
Valdecoxib is also superior to non-selective NSAIDs with regard to ulcer complications of the upper gastrointestinal tract (perforation, bleeding and gastric outlet obstruction). According to a meta-analysis based on data from eight 12–26-week controlled trials in arthritis patients (n = 7434), the annualized incidence (events per total patient-years of exposure) of these events for valdecoxib (0.68%) was approximately three-fold lower than with non-selective NSAIDs (1.96%, P < 0.05), and there was no difference compared with placebo (0%) (Figure 2).43 In the sub-group of patients who were non-aspirin users, the annualized incidence of serious upper gastrointestinal events was even lower (0.29%).
The gastrointestinal safety of members of the ‘coxib’ group given early approval, i.e. celecoxib and rofecoxib, has been examined in two large, randomized, double-blind clinical trials: the Vioxx Gastrointestinal Outcomes Research (VIGOR) trial44 and the Celecoxib Long-term Arthritis Safety Study (CLASS).45 The incidence of serious gastrointestinal events with these two COX-2-selective inhibitors was significantly lower than with traditional NSAIDs, and demonstrated a 50% relative risk reduction in favour of ‘coxibs’. It should be noted, however, that pre-existing risk factors, including advanced age, a history of peptic ulcer and gastrointestinal bleeding, as well as concomitant treatment with aspirin, may limit the advantages of COX-2-selective inhibition over traditional NSAIDs.46,47 Thus, the problem of the co-prescription of low-dose aspirin to reduce cardiovascular risk is a matter of special concern, as the number of cardiovascular patients taking aspirin is growing. The possibility exists that treatment with ‘coxibs’ exposes aspirin users to almost the same risk of severe gastrointestinal complications as seen with non-selective NSAIDs. Whether such patients would benefit from receiving co-treatment with a cytoprotective agent (e.g. misoprostol) or acid inhibitory therapy (e.g. with omeprazole or lansoprazole), as suggested by some authors, remains to be elucidated in future controlled studies.
A comparison of the effect of valdecoxib (40 mg b.d.) and traditional NSAIDs on platelet aggregation has been performed in young and elderly healthy volunteers using single and multiple doses of the drugs.48–50 The effects were evaluated by ex vivo platelet aggregation assays and by measuring the bleeding time immediately following the first dose (day 1) and last dose (day 8).49 Both valdecoxib and placebo had no significant effects on platelet aggregation and bleeding time. In contrast, treatment with diclofenac (75 mg b.d.) and naproxen (500 mg b.d.) resulted in significant inhibition of platelet aggregation and prolongation of the bleeding time due to the inhibition of COX-1 in platelets and a reduction in thromboxane A2 synthesis.51
As COX-2 is constitutively expressed in the kidney and contributes to renal haemodynamics,52 the renal safety of COX-2 inhibition using valdecoxib has been questioned.
Pooled data on renal safety obtained from osteoarthritis and rheumatoid arthritis trials have demonstrated that the incidence of common renal adverse effects (albuminuria, peripheral oedema and hypertension) following selective COX-2 inhibition with 40 mg of valdecoxib is higher (0.9%, 2.3% and 2.8%, respectively) than that for placebo (0.2%, 0.7% and 0.6%, respectively), but does not significantly exceed that for traditional NSAIDs (0.5%, 2.2% and 1.5%, respectively). The majority of renal events are dose dependent and are due either to the occurrence of oedema or an increase in blood pressure. Similar renal adverse effects have also been described for celecoxib53 and rofecoxib,54 raising the question of whether these effects are mediated by COX-2 inhibition in the kidney. In view of these findings and as in treatment with other ‘coxibs’ and traditional NSAIDs, it is recommended that the administration of valdecoxib be accompanied by close monitoring of renal function, especially in subjects in whom renal function is highly dependent on the production of prostaglandins (e.g. in congestive heart failure, cirrhosis and in patients on diuretics).
There are concerns that COX-2-selective inhibitors may be thrombogenic and increase the risk of myocardial infarction. These have largely arisen because of an unexpected high rate of myocardial infarction in patients receiving rofecoxib compared with patients receiving naproxen (VIGOR study),44 although the CLASS trial with celecoxib vs. diclofenac or ibuprofen showed no statistically significant difference between the treatment groups in the incidence of major cardiovascular events.45
The possibility that an increase in cardiovascular risk with ‘coxibs’ is a class effect remains unclear and needs further evaluation. However, the experience with valdecoxib obtained so far shows no evidence of an increased cardiovascular risk with this particular agent. No significant difference in the incidence of major cardiovascular events (events per 100 patient-years) was seen in pooled data from 12–26-week controlled trials in arthritis patients (n = 8752; placebo, 1.2%; valdecoxib 1–80 mg, 1.7%; non-selective NSAIDs, 2.4%). Similarly, there was no significant difference for the corresponding incidences in aspirin users and non-aspirin users (Pharmacia, on file 2000).
Central nervous system effects
COX-2 is expressed constitutively in the central nervous system, including the spinal cord55,56 and various regions in the brain,57 and is involved in the processing of special sensory input and in the elaboration of autonomic, endocrine and behavioural responses.58,59 These findings raise the question of whether the inhibition of COX-2 in the central nervous system with ‘coxibs’ and NSAIDs may be responsible for the common drug-related central nervous system disorders, such as dizziness and headache, as well as the less common reactions, such as somnolence, vertigo and paraesthesia.
In less than a decade since the discovery of COX-2, COX-2-selective inhibitors have achieved a firm role in the treatment of osteoarthritis, rheumatoid arthritis and the management of acute pain. Valdecoxib, a new inhibitor of COX-2, exhibits the advantages of this class of anti-inflammatory agents in terms of less gastrointestinal toxicity and lack of an effect on platelets, whilst having an efficacy comparable with that of traditional NSAIDs. However, the incidence of adverse effects on the kidney (albuminuria, peripheral oedema and hypertension) is comparable with that of classical NSAIDs. Further clinical experience with valdecoxib is necessary in order to fully explore the therapeutic potential and tolerability of selective inhibitors of COX-2, not only in the treatment of inflammatory conditions and pain management, but also in novel indications, including the treatment of cancer and neurological diseases.
The authors thank Dr Ruth Fricker (Pharmacia, Germany) for provision of direct access to the original data files on valdecoxib.