The safety profile, tolerability, and effective dose range of rofecoxib in the treatment of rheumatoid arthritis

CLINICAL

THERAPEUTICYIVOL.

2 I, NO. 10, 1999

The Safety Profile, Tolerability, and Effective Dose Range of Rofecoxib in the Treatment of Rheumatoid Arthritis Thomas J. Schnitzer, MD, PhD,’ Ken Truit$ MD: Roy Fleischmann, MD: Paul Dalgin, MD,“ Joel Block, MD: Qi Zeng, PhDt James Bolognese, MS&t: Beth Seidenberg, MD: and Elliot W. Ehrich, MD; for the Phase II Rofecoxib Rheumatoid Arthritis Study Group ‘Qfice of Clinical Research and Training, Northwestern University School of Medicine, Chicago, Illinois, 2Merck Reseurch Laboratories, Rahway, New Jersey, 3A4etroplex Clinical Research Center, Dallas, Texas, 4Stamford Therapeutics Consortium, Stumford, Connecticut, and 5Rush-Presbyterian St. Luke :r Medical Center, Chicago, Illinois

ABSTRACT Nonsteroidal anti-inflammatory drugs (NSAIDs) inhibit both cyclooxygenase (COX)-1 and COX-2 isoenzymes and are effective in the treatment of inflammatory disorders. This &week, double-masked, placebo-controlled trial was undertaken to assess the safety profile, tolerability, and effective dose range of once-daily rofecoxib, a COX-2-specific inhibitor, in the treatment of rheumatoid arthritis (RA). After a 3- to 15-day washout of prior NSAID therapy, 658 patients were randomly allocated to receive placebo or rofecoxib 5 mg, 25 mg, or 50 mg once daily. Safety pro file. tolerability, and efficacy were evaluated after 2,4, and 8 weeks of therapy. Six hundred fifty-eight patients (168, 158,17 I, and 161 in the placebo and 5-mg, 25-mg, ‘Deceased. Accepted for publication August 3, 1999. Printed in the USA. Reproduction in whole or part is not permitted.

1688

and 50-mg rofecoxib groups, respectively) were enrolled at 79 clinical centers in the United States. Mean age was 55 years, mean duration of RA was IO years, and 506 (77%) of the 658 patients were female. All groups had similar baseline demographic characteristics. Patients taking rofecoxib 25 and 50 mg showed significant clinical improvement compared with those taking placebo; 43.9% in the rofecoxib 25-mg group and 49.7% in the rofecoxib 50-mg group completed the treatment period and achieved an American College of Rheumatology 20 response (P = 0.025 and 0.001 vs placebo, respectively). The 5-mg dose of rofecoxib did not differ significantly from placebo. Patients in the rofecoxib 25- and 50-mg groups showed significant improvement in key individual efficacy measurements, including patient global assessment of pain, patient and investigator global assessmentof disease activity, and Stanford Health Assessment Questionnaire Disability lndex (P < 0.05 vs placebo). Compared with placebo, sig0149.2918/99/S

I9.00

T.J. SCHNITZER ET AL.

nificantly fewer patients in the 25mg and 50-mg rofecoxib groups discontinued therapy because of lack of efficacy (P = 0.02 and P = 0.032, respectively). Our results show that rofecoxib 25 and 50 mg once daily was effective and generally welltolerated in patients with RA. Key words: rofecoxib, cyclooxygenase, COX-2-specific inhibitors. rheumatoid arthritis. INTRODUCTION Nonsteroidal anti-inflammatory drugs (NSAIDs) are used to treat pain and inflammation caused by a variety of clinical disorders. In patients with rheumatoid arthritis (RA), NSAIDs reduce joint pain and tenderness and thereby improve function. 1 However, chronic use of NSAIDs is associated with gastrointestinal (GI) toxicity, which has limited their use and led to the search for safer alternative agents. Cyclooxygenase (COX) catalyzes the production of prostaglandins (PCs) from membrane-associated arachidonic acid, and COX inhibition underlies both the therapeutic and toxic effects of traditional NSAIDS.~,” Only 2 distinct COX isoforms (COX- 1 and COX-2) have been identified and well characterized.4-7 They share a 60% amino acid homology,* similar tertiary structures,’ and similar but nonidentical active sites.“) The constitutive expression and wide distribution of COX- 1‘I suggest that COX- 1-derived PGs mediate routine physiologic functions, including gastric mucosal protection12 and vascular hemostasis.t3 In contrast, COX-2 is principally an inducible enzyme that is constitutively expressed in certain specialized tissues, including brain, testes, ovary, and kidney. ” Inflammation and other stressors can markedly increase COX-2 expression.‘3--19 Sites of inducible COX-2

expression include rheumatoid synoviocytes,1’~20macrophages,11*21and polymorphonuclear leukocytes.“g2’ NSAIDs inhibit both COX- 1 and COX-2. COX-2 inhibition principally underlies the analgesic and anti-inflammatory effects of NSAIDS.~~,~~In contrast, the GI toxicity of NSAIDs primarily results from the nontherapeutic inhibition of COX- 1.24-28 Agents that specifically inhibit COX-2 would provide an improved therapeutic index (in particular, improved GI safety) compared with traditional NSAIDS.~‘,~~ Rofecoxib* (MK-0966; 4-[4-(methylsulfonyl)phenyl]-3-phenyl-2(5H)-furanone) specifically inhibits COX-2. Ex vivo human whole blood assays showed dose-related inhibition of COX-2 activity without detectable inhibition of COX- 1 activity after single oral doses (5 to 1000 mg)30 and multiple oral doses (25 to 375 mg/d for 14 days).“’ Once-daily doses of 12.5 mg and 2.5 mg of rofecoxib are effective in the treatment of patients with osteoarthritis (OA),32,33 and 50 mg is effective in control of acute pain.34s3s Hence therapeutic effects of rofecoxib have been seen at doses that lack COX-1 inhibition as detected by currently available assays.“O In patients with RA, rofecoxib may be as effective as traditional NSAIDs but with improved GI safety. Patients with RA are particularly prone to develop NSAID-associated GI toxicity based on several risk factors, most notably a continuous high-dose exposure to NSAIDs and a high rate of concomitant corticosteroid use.’ The current double-masked, placebo-controlled trial of 3 doses of rofecoxib was undertaken to determine *Trademark: Vioxx@ (Merck & Co.. Inc., West Point, Pennsylvania).

1689

CLINICAL

the drug’s optimal dose range, safety, and tolerability in patients with RA.

PATIENTS AND METHODS Patients Men and women 218 years of age and within 40% of ideal body weight based on 1983 Metropolitan Life Insurance Tables’6 whose diagnosis of RA met the 1987 American College of Rheumatology (ACR) criteria were eligible.“7 A diagnosis of RA must have been made after age 16 and 26 months before study enrollment. Patients were required to have a history of therapeutic benefit from NSAIDs and to have used therapeutic doses for 225 of the 30 days preceding study entry. Stable therapy (for the previous 6 months) with certain antirheumatic drugs (oral or intramuscular gold salts, sulfasalazine, azathioprine, D-penicillamine, or hydroxychloroquine) was permitted at study entry. Up to one third of patients were allowed to continue taking low-dose oral corticosteroids (57.5 mg prednisone/d or equivalent) if use had been continuous in the past 3 months and the dose had been stable for the past 30 days. Up to one third of patients were permitted to take methotrexate, provided that the weekly dose was 220 mg and had been stable for the past 3 months. This was enforced by capped enrollment. Before randomization, patients taking both steroids and methotrexate were excluded. Patients were excluded if they had systemic lupus erythematosus, spondyloarthropathy, polymyalgia rheumatica, gout, Paget’s disease, active GI bleeding or ulceration, a positive stool guaiac screen, or uncontrolled diabetes. Other exclusion cri1690

THERAPEUTICS@

teria were a recent serious cardiovascular event (eg, myocardial infarction in the past year), active hepatitis, malignancy, serum creatinine >2.0 mg/dL, or estimated creatinine clearance 130 mL/min, serum aminotransferase levels 2150% of the upper limit of normal, allergy to acetaminophen, aspirin, or NSAIDs, or a history of bronchoconstriction associated with nasal polyps. Study Design The study was approved by the institutional review boards at all participating centers. After giving written informed consent, potential participants underwent a screening assessment and standardized assessment of RA severity. Eligible patients underwent a prestudy NSAID washout period (3 to 15 days) based on the elimination half-life of the discontinued agent. After washout, patients returned for reassessment and were randomly assigned to the active treatment protocol if they satisfied the following disease activity and disease flare criteria: (1) a score of 240 mm on the patient global assessment of disease activity ( IOO-mm visual analog scale [VAS]), with an increase from the screening visit of 215 mm; (2) 29 tender joints, with an increase of 220% from the screening visit; and (3) 26 swollen joints. Patients were randomly assigned to 1 of 4 study arms: placebo, rofecoxib 5 mg/d, rofecoxib 25 mg/d, or rofecoxib 50 mg/d. Assignment to treatment was stratified based on concomitant methotrexate use. Concomitant corticosteroid use was documented but was not used in a formal stratification scheme. Follow-up assessments were performed after 2,4, and 8 weeks of study treatment or at discontinuation.

T.J. SCHNITZER ET AL.

Dose Selection and Pain Management

Safety Assessments

Dose selection was based on results from prior clinical studies with rofecoxib32”3; doses of 12.5 to 25 mg/d have demonstrated efficacy in the treatment of OA. Because RA is associated with a greater and more consistent degree of joint inflammation than is OA, the doses in the present study included the maximum effective OA dose (ie, 25 mg/d) as well as a 50-mg/d dose. Acetaminophen (up to eight 325mg tablets/d with a maximum dose of 24 tablets in a 2-week period) was provided to patients throughout the study as rescue therapy for breakthrough pain. Patients were encouraged to use as little rescue therapy as possible.

Adverse effects were recorded at each visit. All adverse effects were reviewed by the investigator and rated by severity and relation to study treatment. Investigators were masked to treatment. Physical examination was performed and vital signs, including weight and stool guaiac screen, were obtained at each visit. Clinical laboratory assessments (complete blood count [with differential and platelet count], a blood chemistry panel [electrolytes, calcium, albumin, total protein, aspartate aminotransferase, alanine aminotransferase (ALT), alkaline phosphatase, total bilirubin, blood urea nitrogen, creatinine, glucose, and uric acid], and a dipstick urinalysis [for protein, glucose, pH, and red and white blood cell counts]) were obtained at each visit. All testing was performed at a central laboratoq-(Medical Research Laboratories, Highland Heights, Kentucky).

Efficacy End Points Treatment response was assessedby the number of patients in each group who met the criteria for an ACR 20 response (20% improvement in tender and swollen joint counts and 20% improvement in 3 of 5 remaining ACR core measures).jX Individual ACR core efficacy measurements used as end points in this trial were (1) swollen joint count (66 joints), (2) tender joint count (68 joints), (3) patient global assessment of disease activity on a IOOmm VAS ranging from 0 (very well) to 100 (very poor), (4) investigator global assessmentof disease activity on a 5-point Likert scale ranging from 0 (very well) to 4 (very poor), (5) Stanford Health Assessment Questionnaire (HAQ) Disability Index on a 4-point Likert scale ranging from 0 (without any difficulty) to 3 (unable to do), (6) patient global assessment of pain on a IOO-mm VAS ranging from 0 (no pain) to 100 (extreme pain), and (7) C-reactive protein level .j9

Statistical Analysis The number of patients satisfying the ACR 20 criteria,36 discontinuing treatment, and experiencing adverse events or abnormal laboratory values were individually compared between treatment groups using the Fisher exact test. For each effect variable included in the composite index, the average change from baseline across the 8week treatment period was compared for the 5-, 25-, and 50-mg rofecoxib doses versus placebo. End points were analyzed by an analysis of variance model with terms for study center, stratum, baseline covariate, and treatment. The mean change from baseline (284% confidence intervals [CIs]) at each week was plotted for each efficacy variable. Nonoverlap1691

CLINICAL

THERAPEUTICS”

tivity criteria after NSATD washout, 35 (19.6%) did not give written informed consent, 33 (18.4%) had significant prestudy clinical or laboratory abnormalities, and 22 (12.3%) had changed their medical therapy in the past month. The remaining 25 patients failed screening based on various inclusion and exclusion criteria. Baseline characteristics were similar among treatment groups; there were no statistically significant differences (Table I). Overall, 545 (82.8%) patients completed the trial. Significantly fewer patients

ping 84% CIs indicated a statistically significant difference (P 5 0.0.5).4”

RESULTS Of 837 patients screened, 658 (78.6%) were enrolled and randomly assigned to treatment at 79 clinical centers in the United States. One hundred seventy-nine patients were screened but not randomized. The most common reasons for screening failure were as follows: 64 (35.8%) failed to meet either flare or acTable I. Patient characteristics at baseline.

Rofecoxib Variable

25 mg (n = 171)

50 mg (n = 161)

Overall (n = 658)

135 (78.9) 36 (21.1) 55.7 26-8 I 78 49-141 9 &48

130 (80.7) 31 (19.3) 54.4 27-76 74 42-130 10 O-61

506 (76.9) 152 (23.1) 54.9 24-86 76 42-153 10 O-61

23 (14.6) 106 (67.1) 29 (18.4)

35 (20.5) 10.5(61.4) 31 (18.1)

35 (21.7) 98 (60.9) 28 (17.4)

127 (19.3) 420 (63.8) I I1 (16.9)

39 (24.7) 39 (24.7) 15 (9.5) 16 (10.1) 5 (3.2) 4 (2.5) I (0.6)

58 (33.9) 41 (24.0) 18 (10.5) 7 (4.1) 5 (2.9) 2 (1.2) 0 (0)

41 (25.5) 33 (20.5) 14 (8.7) 10 (6.2) 7 (4.3) 3 (1.9) 1 (0.6)

189 (28.7) 137 (20.8) 60 (9.1) 47 (7.1) 26 (4.0) 13 (2.0) 6 (0.9)

36 (22.8)

35 (20.5)

36 (22.4)

145 (22.0)

Placebo

5 mg

(n = 168)

(n = 158)

Sex, no. (%) Female 121 (72.0) Male 47 (28.0) 54.7 Mean age, y 24-86 Range 76 Mean body weight, kg 45-131 Range Mean duration of RA, y 8 Range O-54 ARA functional class, no. (%) I 34 (20.2) 11 111 (66.1) III 23 (13.7) Use of DMARDs, no. (%) Methotrexate 51 (30.4) Antimalarials 24 (14.3) Sulfasalazine 13 (7.7) 14 (8.3) Gold (oral and injectable) Azathioprine 9 (5.3) Tetracycline and derivatives 4 (2.4) Penicillamine 4 (2.4) Use of systemic corticosteroids, no. (%) 38 (22.6)

120 (75.9) 38 (24.1) 54.8 26-80 76 44-153 II O-6 1

RA = rheumatoid arthritis; ARA = American Rheumatology antirheumatic drugs.

1692

Association;

-

DMARDs

= disease-modifying

T.J. SCHNITZER ET AL.

Table 11. Number (%) of patients completing or discontinuing the trial. Rofecoxib 5 mg (n = 158)

25 mg (n = 171)

50 mg (n = 161)

131 (78.0) 37 (22.0)

134(84.8) 24 (15.2)

145(84.8) 26(15.2)

135(83.9) 26 (16.1)

5 (3.0) 0 (0)

5(3.2) 1 (0.6) 16(10.1) 2 (1.3)

Placebo

(n = 168) Completed Discontinued Clinical adverse experiencesor events Abnormal laboratory values Lack of efficacy Othert

24(14.3) 8 (4.8)

8 (4.7) 1 (0.6)

11 (6.4)* 6(3.5)

*P = 0.020 versusplacebo. ‘P = 0.032 versusplacebo. kludes patientslost to follow-up, deviations from protocol. and patient discontinuation/withdrawal in the 25.mg and 50-mg rofecoxib groups than in the placebo group discontinued therapy because of lack of efficacy (P = 0.020 and P = 0.032, respectively) (Table II). Efficacy Table III shows the number of patients who achieved an ACR 20 response. In the rofecoxib 25mg group, 74 patients (43.9%) achieved an ACR 20 response and completed the study (P = 0.025 vs placebo). Similarly, in the rofecoxib 50mg group, 80 patients (49.7%) achieved an ACR 20 response and completed the study (P = 0.001 vs placebo). The 5-mg dose of rofecoxib did not differ significantly from that of placebo for patients who completed the study (33.5% vs 3 1.7%, respectively). Rofecoxib demonstrated similar efficacy in methotrexate users and nonusers. In addition to significant responsesto the composite ACR 20 response index, patients in the rofecoxib 25-mg and 50-mg groups

10 2 1I 3

(6.2) (1.2) (6.8)+ (1.9)

of consent.

showed significant improvement in key individual efficacy measurements compared with the placebo group (patient global assessment of pain [Figure I], patient global assessment of disease activity [Figure 21, investigator global assessment of disease activity [Figure 31, and Stanford HAQ Disability Index score [Figure 4)). No significant treatment effect with either rofecoxib or placebo was seen for C-reactive protein. The use of acetaminophen rescue also was compared between groups. From baseline to week 8, the 25mg and 50-mg rofecoxib groups used significantly less acetaminophen than did the placebo group (P = 0.009 and P < 0.001, respectively). Safety Profile and Tolerability Safety results are reported for the g-week treatment period. No significant betweengroup differences were observed in the number of patients who discontinued treatment because of adverse effects (Table II). Serious adverse effects occurred in 12 (1 .S%) of the 658 random-

1693

CLINICAL THERAPEUTICS”

Table III. Number (%) of patients who achieved a 20% improvement in tender and swollen joint counts and a 20% improvement in 3 of 5 remaining measures of the American College of Rheumatology (ACR 20). Rofecoxib

Total population (N = 658) ACR 20 responders* ACR 20 responders and completers* (%) Methotrexate users (n = 192) ACR 20 responders Methotrexate nonusers (n = 163) ACR 20 responders

25 mg

50 mg

Placebo

5 mg

58/l 68 (34.7)

56/158 (35.4)

82/l 7 I (48.5)+

86/161 (53.4)+

53/168 (3 I .7)

53/l% (33.5)

74/171 (43.9)+

80/161 (49.7)+

17/50 (34.0)

14/38 (36.8)

24/54 (44.4)

26/50 (52.0)

36/l 17 (30.7)

39/l 20 (32.5)

50/115 (43.5)’

54/l I I (48.6)s

20 respondersrepresent the number (%) of patients who achieved an ACR 20 responseregardless of whether they completed the study. ACR 20 respondersand completers represent the number (%) of patients who achieved an ACR 20 responseand completed the R-week study.

*ACR

‘1 = 0.001 versus placebo. *P = 0.057 versus placebo. $P = 0.007 versus placebo.

5 -3

P Placebo A Rofecoxib 41 Rofecoxib o Rofecoxib

O-

-5 -

2

5 mg 25 mg 50 mg

0

3

h

-lO-

UP a $?

-15-

‘2 .E r

-20 -

$.g

-25

2 E -3()VE 2 .c -35r”

40-45 -50

I R

I

I 2

I

I

I

I

I

1 8

Time (I&

Figure 1. Mean change in patient global assessmentof pain from baseline. Score is based on a 100~mmvisual analog scale (VAS) ranging from 0 (no pain) to 100 (extreme pain). P < 0.001 for 25-mg and 50-mg groups versus placebo. R = randomization. 1694

T.J. SCHNITZER ET AL. P A * o

5-

2 ‘05 :: -

Placebo Rofecoxib Rofecoxib Rofecoxib

5 mg 25 mg 50 mg

-5-

-IO-

2 $? -15=-$ .E r -2oE.5

-25-

2 E -3OoE g .E -352

-4o-45-50

’

I

I

I

I

1

I

I

I

I

8

2

R

Time (ik)

Figure 2. Mean change in patient global assessment of disease activity from baseline. Score is based on a IOO-mm visual analog scale (VAS) ranging from 0 (very well) to 100 (very poor). P < 0.001 for 25mg and 50-mg groups versus placebo. R = randomization. P A # o

0.0 -

-0.2 2 8

Placebo Rofecoxib Rofecoxib Rofecoxib

5 mg 25 mg 50 mg

-0.4 -

co r al” 3 u

-0.8 -

‘(u cg co

-1 .o -

Pg gf 5

-1.2

-

C

-1.4

-

fz 3

-1.6

-

-1.8

-

-0.6 -

0

-2.0

I R

I

I

2

I

I

4

I

I

I

1

8

Time (wk)

Figure 3. Mean change in investigator global assessmentof disease activity from baseline. Score based on a 5-point Likert scale ranging from 0 (very well) to 4 (very poor). P < 0.001 for 25-mg and 50-mg groups versus placebo. R = randomization. 1695

CLINICAL THERAPEUTICS” P Placebo A Rofecoxib 5 mg 4k Rofecoxib 25 mg 0 Rofecoxib 50 mg

0.00c?

-0.08

-

-0.18

-

-0.28

-

-0.38

-

8 cn g

5

.c

8 *

aa or 22. = .= 0 5 f

-0.48

’

I R

1

1 2

I

I 4

I

I

I

I 8

Time (wk) Figure 4. Mean change in Stanford Health Assessment Questionnaire Disability Index from baseline. Score is based on a 4-point Likert scale ranging from 0 (without any difficulty) to 3 (unable to do). P < 0.001 for 25mg and 50-mg groups versus placebo. R = randomization. ized patients. Only 1 adverse effect-infectious gastroenteritis in a patient in the rofecoxib 25-mg group-was considered possibly related to the study drug. Only 4 patients (0.6%) discontinued therapy during the g-week treatment period (elevated ALT. 1 patient in the S-mg rofecoxib group; elevated serum creatinine. I patient each in the 2S-mg and 50-mg rofecoxib groups; and decreased hematocrit, 1 patient in the SO-mg group) because of abnormal laboratory values. Table IV summarizes clinical adverse effects that occurred in >3% of patients in any group. The most common adverse effects were diarrhea, headache, and fatigue. Although 7 patients in the 25mg group reported dizziness, no dose-related trend was observed. No significant between-

group differences were noted. More patients in the rofecoxib SO-mg group than in the other 3 groups experienced a rash. Most rashes were mild, were considered by the investigator to be unrelated to study drug, and resolved during treatment. No rashes were associated with angioedema or anaphylaxis, and none was considered a serious adverse effect. Only 2 patients, 1 in the rofecoxib 5-mg group and I in the rofecoxib 50-mg group, discontinued therapy because of a rash. Because NSAIDs have known renal and vascular effects, adverse experiences potentially indicative of fluid retention or blood pressure elevation were examined in detail. The incidence of lower-extremity edema was similar in all 4 groups (1.20/o, 1.3%, 2.3%, and 2.5% in the placebo and

T.J. SCHNITZER ET AL

Table IV. Adverse effects (%) occurring in >3% of any treatment group.* Rofecoxib 50 mg

Adverse Effect

(n = 168)

(n = 1.58)

25 mg (n = 171)

(n = 161)

Abdominal pain AsthenWfatigue

1 (0.6) 8 (4.8) 5 (3.0) 3 (1.8) 3 (1.8) 11 (6.5) 2 (1.2) 7 (4.2) 2 (1.2) 6 (3.6)

3 (1.9) 5 (3.2) 1I (7.0) 1 (0.6) 4 (2.5) 7 (4.4) 0 (0) 4 (2.5) 4 (2.5) 4 (2.5)

1 (0.6) 7 (4.1) 8 (4.7) 12 (7.0) 1 (0.6) 10 (5.8) 5 (2.9) 6 (3.5) 3 (1.8) 5 (2.9)

6 (3.7) 4 (2.5) 6 (3.7) 6 (3.7) 5 (3.1) 7 (4.3) 5 (3.1) 5 (3.1) 9 (5.6) 6 (3.7)

Placebo

Diarrhea Dizziness

Dyspepsia Headache

Hypertension Nausea Rash Sinusitis

5w

“Incidence is ~3% in >l treatment group. Regardlessof the number of adverse effects per body system, each patient is countedonly once per body system.

rofecoxib 5-mg, 25mg, and 50-mg groups, respectively). Five patients discontinued treatment because of edema (2 in the placebo group, 1 in the rofecoxib 25mg group, and 2 in the rofecoxib 50-mg group). The incidence of hypertension was similar among the groups (1.2%, O%, 2.9%, and 3. I % in the placebo and rofecoxib S-mg, 25-mg, and 50-mg groups, respectively). No patients discontinued treatment because of increased blood pressure. No significant between-group differences were observed in the percentage of patients who discontinued treatment because of GI adverse effects (O%, 1.3%, 1.8%, and 0% in the placebo and rofecoxib 5-mg, 25mg, and 50-mg groups, respectively). No gastroduodenal perforations, ulcers, or bleeds were reported during the study. No meaningful differences in adverse effects or laboratory values were seen in methotrexate users and nonusers.

DISCUSSION The present study demonstrated that rofecoxib, a COX-2-specific inhibitor, was generally well-tolerated and was associated with significant improvement in symptoms of RA. Previous studies have demonstrated its efficacy in OA.32z33In these studies, rofecoxib showed clinical efficacy similar to high doses of traditional NSAIDs. In this phase II study, rofecoxib 2.5mg/d and 50 mg/d produced statistically and clinically significant efficacy in the treatment of RA. The 5-mg dose did not differ significantly from placebo. The differences between the 25- and 50-mg rofecoxib groups were not statistically significant. Clinical efficacy was evident at the first assessment (week 2) and was sustained throughout the g-week trial. The number of patients who met the criteria for an ACR 20 response was significantly 1697

CLINICALTHERAPEUTICS" greater in the 25 and 50-mg rofecoxib groups than in the placebo group. The average treatment response for most of the individual outcome measures included in the ACR composite index was also significantly greater in the 25- and 50-mg rofecoxib groups than in the placebo group. Consistent with many clinical trials in patients with RA, a substantial placebo effect was observed in the present study.41A2Approximately one third of patients in both the placebo and 5-mg rofecoxib groups met ACR 20 criteria.39*N’ Trial design, patient selection, and patients’ beliefs and expectations are recognized factors contributing to the placebo effect in RA trials.sgAO The clinical efficacy of rofecoxib was clearly demonstrated against a background of stable disease-modifying antirheumatic drugs, particularly methotrexate. In subgroup analyses, rofecoxib demonstrated similar safety and efficacy in methotrexate users and nonusers. Prior pharmacokinetic studies have shown that high doses (75 mg) of rofecoxib caused a 23% increase in methotrexate area under the curve (AUC) 0 to 24 hours after single-dose administration but did not prolong plasma exposure as assessed by methotrexate levels 24 hours after dosing (B. Gertz, MD, PhD, unpublished data, 1998). Traditional NSAIDs, particularly salicylates, interact with methotrexate. Concomitant use of NSAIDs and methotrexate can increase methotrexate concentrations43.ti and decrease methotrexate clearance.‘“;?6 The risk of methotrexate toxicity appears to increase with prolongation of plasma methotrexate levels rather than with changes in maximum concentration or AUC per se.47 It is therefore unlikely that the effects on plasma methotrexate levels observed with rofecoxib could rep1698

resent a toxicity problem in patients with RA, as rofecoxib did not prolong plasma methotrexate exposure. In the current study, no adjustments in methotrexate dose were required, and no safety problems with methotrexate treatment were noted. The results of this study suggest that standard ACR guidelines for monitoring methotrexate therapy should be used in patients receiving methotrexate and rofecoxib.48 Rofecoxib demonstrated excellent tolerability in patients with RA. There were no reported cases of gastroduodenal perforation, ulcers, or bleeds during this study. In chronic users, NSAID-induced gastropathy can result in substantial morbidity and mortality, including a 2% to 4% annual incidence of perforation, ulcers, or bleeds. The GI profile of rofecoxib is consistent with prior reports of its superior GI safety and tolerability compared with NSAIDs in patients with OA!9Jo Long-term studies directly comparing rofecoxib with NSAlDs are necessary to definitively demonstrate improved GI safety in patients with RA. Fluid retention, the most common NSAID-related renal effect, is clinically detectable in -5% of patients receiving NSAIDs.” In this study, no clinically meaningful effects on mean blood pressure or edema were associated with rofecoxib treatment. The rates of edema observed in the 25-mg and 50-mg rofecoxib groups were well within the range seen with traditional NSAIDs. CONCLUSIONS Treatment with rofecoxib, which specifically inhibits COX-2, was well tolerated and produced clinically significant improvements in the signs and symptoms of RA. These results show that specific inhi-

T.J. SCHNITZER ET AL.

bition of COX-2 (without inhibition of COX- 1) can produce clinically meaningful improvements in RA symptoms and that COX-2-derived PGs are important mediators in RA-associated pain and inflammation. Ongoing long-term studies of rofecoxib in the treatment of RA will define the optimal dose, durability of efficacy, tolerability, safety, and comparability to traditional NSAIDs and other COX-2-specific inhibitors.

ACKNOWLEDGMENTS Study investigators were as follows: Kenneth Bahrt, South Plainfield, New Jersey; Arthur Bankhurst, Albuquerque, New Mexico; Herbert Baraf, Wheaton, Maryland; Joel Block, Chicago, Illinois; Barry Bockow, Seattle, Washington; Peter Bonafede, Portland, Oregon; Robert Bonebrake, Madison, Wisconsin; A. McKay Brabham, Columbia, South Carolina; Jacques Caldwell, Gainesville, Florida; Julian Colton, St. Petersburg, Florida; John Conte, Barrington, Rhode Island; William S. Cryan, Long Beach, California; Paul Dalgin, Stamford, Connecticut; Robin K. Dore, Anaheim, California; William M. Edwards, North Charleston, South Carolina; H. Walter Emori, Medford, Oregon; John E. Ervin, Kansas City, Missouri; Roy M. Fleischmann, Dallas, Texas; Michael Franklin, Willow Grove, Pennsylvania; Marc Goldberg, Passaic, New Jersey; AlIan Goldman, Milwaukee, Wisconsin; Gary Gordon, Wynnewood, Pennsylvania; William Gough,Asheville, North Carolina; William B. Gruhn, Charlotte, North Carolina; John M. Hague, Indianapolis, Jndiana; Matthew D. Heller, Peabody, Massachusetts; Peter A. Holt, Baltimore, Maryland; Jeffrey Kaine, Sarasota, Florida; James D. Keamey. Camp Hill, Pennsylva-

nia; Alastair Kennedy, Melbourne, Florida; Dan Kirby, Melbourne, Florida; Alan Kivitz, Altoona, Pennsylvania; Steven J. Klein, Cumberland, Maryland; Robert Levin, Palm Harbor, Florida; Donald Levitt, Peoria, Arizona; Charles Ludivico, Pennsylvania; Michael Bethlehem, Luggen, Cincinnati, Ohio; Raymond L. Malamet, Hagerstown, Maryland; David Mandel, Mayfield Village, Ohio; Joseph A. Markenson, New York, New York; Lawrence McAdam, Thousand Oaks, California; Harris McIlwain, Tampa, Florida; James McKay, Tulsa, Oklahoma; George McLaughlin, Norristown, Pennsylvania; Phillip Mease, Seattle, Washington; Stephen Miller, Las Vegas, Nevada; Robert A. Moidel, Sellersville, Pennsylvania; David Neustadt, Louisville, Kentucky; C. Michael Neuwelt, San Leandro, California; James H. Newman, Wilmington, Delaware; Richard Newton, Wenatchee, Washington; Kenneth M. Nies, Torrance, California; Michael Peveler, Louisville, Kentucky; Jeffrey Poiley, Orlando, Florida; Niall Roche, Fremont, California; Philippe Saxe, Delray Beach, Florida; Michael Schiff, Denver, Colorado; Margo Schilling, Iowa City, Iowa; Thomas J. Schnitzer, Chicago, Illinois; Michael Schweitz, West Palm Beach, Florida; James Seibold, New Brunswick, New Jersey; Stephen R. Shaul, Yakima, Washington; William Shergy, Huntsville, Alabama; Sheldon Solomon, Cherry Hill, New Jersey; Peter Stein, Santa Rosa, California; Steven Stem, Louisville, Kentucky; John R.P. Tesser, Phoenix, Arizona; Robert G. Trapp, Springfield, Illinois; Michael J. Tutt, Tucson, Arizona; Daniel Wallace, Los Angeles, California; Nathan Wei, Frederick, Maryland; Richard H. White, Sacramento, California; Craig W. Wiesenhutter, Coeur d’Alene, Idaho; Christopher M. Wise, Richmond, Virginia. 1699

CLINICAL

In addition we would like to acknowl-

edge the administrative and scientific efforts of Rhoda Sperling, MD, Patricia Brady, Rashmi Jain, Earl St. Rose, Michele Fabriele, Marjorie Anderson, Lisa De Tora, and Joseph Giuliano.

Address correspondence to: Ken Truitt, MD, Merck Research Laboratories, RY 32-645, 126 East Lincoln Avenue, Rahway, NJ 07065. REFERENCES

THERAPEUTICS”

7. O’Banion MK, Winn VD, Young DA. cDNA cloning and functional activity of a glucocorticoid-regulated inflammatory cyclooxygenase. Proc Nat1 Acad Sci USA. 1992;89:4888-4892. 8. Jones DA, Carlton DP, McIntyre TM, et al. Molecular cloning of human prostaglandin endopemxide synthase type II and demonstration of expression in response to cytokines. J Biol Chem. 1993;268:9049-9054. 9. Luong C, Miller A, Bamett J, et al. Flexibility of the NSAID binding site in the structure of human cyclooxygenase-2. Nur Struct Biol. 1996; I 1:927-933.

I. ACR Clinical Guidelines Committee. Guidelines for the management of rheumatoid arthritis. Arthritis Rheum. 1996;39:7 13-722.

IO. Kurumbail RG, Stevens AM, Gierse JK, et al. Structural basis for selective inhibition of cyclooxygenase-2 by anti-inflammatory agents. Nature. 1996;384:644-647.

2. Vane JR. Inhibition of prostaglandin synthesis as a mechanism of action of aspirin-like drugs. Nut New Biol. 1971;231:232-235.

II

3. Smith JB, Willis AI.. Aspirin selectivity inhibits prostaglandin production in human platelets. Nut New Biol. 1971;231: 235-237.

12 Whittle BJR, Higgs GA, Eakins KE, et al. Selective inhibition of prostaglandin production in inflammatory exudates and gastric mucosa. Nature. 1980;284:27 l-273.

4. Sirios J, Simmons DL, Richards J. Hormonal regulation of messenger ribonucleic acid encoding a novel isoform of prostaglandin endoperoxide H synthase in rat preovulatory follicles. Induction in vivo and in vitro. J Biol Chem. 1992;267: 11586-l 1592.

13 Moncada S, Gryglewski R. Bunting S, Vane JR. An enzyme isolated from arteries transforms prostaglandin endoperoxides to an unstable substance that inhibits platelet aggregation. Nature. 1976;263:663-66.5.

5. Kujubu DA, Fletcher BS, Varnum BC, et al. TIS 10, a phorbol ester tumor promoterinducible mRNA from Swiss 3T3 cells, encodes a novel prostaglandin synthaseicyclooxygenase homologue. J Hiol C’hem. 199 I ;266:286&2872. 6. Xie WL, Chipman JG, Robertson DL, et al. Expression of a mitogen-responsive gene encoding prostaglandin synthase is regulated by mRNA splicing. Proc Nat1 Acud Sci USA. 1991;88:2692-2696. 1700

Vane JR, Botting RM. Anti-inflammatory drugs and their mechanism of action. Inj2amm Res. 1998;47(Suppl 2):578-S87.

14 Walenga RW, Kester M, Coroneos E, et al. Constitutive expression of prostaglandin endopemxide GM synthetase(PGHS)-2 but not PGHS-1 in human tracheal epithelial cells in vitro. Prost#andins. 1996;52:341-359. 15. Akarasereenont P, Mitchell JA, Appleton I, et al. Involvement of tyrosine kinase in the induction of cycle-oxygenase and nitric oxide synthase by endotoxin in cultured cells. Br .I Pharmacol. 1994;113: 1522-1528. 16. Blanc0 A, Habib A, Levy-‘lbledano S, Maclouf J. Involvement of tyrosine ki-

T.J. SCHNITZER ET AL.

nases in the induction of cyclo-oxygenase2 in human endothelial cells. Biochem J. 1995;312:419-423. 17. Lyons-Giordano B, Pratta MA, Galbraith W, et al. Interleukin-1 differentially modulates chondrocyte expression of cyclooxygenase-2 and phospholipase A2. Exp Cell Res. 1993;206:58-62. 18. Mitchell JA, Akarasereenont P, Thiemermann C, et al. Selectivity of nonsteroidal anti-inflammatory drugs as inhibitors of constitutive and inducible cycle-oxygenase. Proc Nut1 Acud Sci USA. 1994; 90: 1169311697. 19. Berenbaum F, Jacques C, Thomas G, et al. Synergistic effect of interleukin-1 B and tumor necrosis factor a or PGE2 production by articular chondrocytes does not involve PLA2 stimulation. Exp Cell Res. 1996;222:379-384. 20. Siegle 1, Klein R, Backman JT, et al. Expression of cyclooxygenase 1 and cyclooxygenase 2 in human synovial tissue: Differential elevalion of cyclooxygenase 2 in inflammatory joint diseases. Arthritis Rheum. 1998;41:122-129. 21. Seibert K, Masferrer JL. Role of inducible cyclooxygenase (COX-2) in inflammation. Receptor: 1994;4: 17-23. 22. Meade EA, Smith WL, deWitt DL. Differential inhibition of prostaglandin endoperoxide synthase (cyclooxygenase) isozymes by aspirin and other nonsteroidal anti-inflammatory drugs. .I Biol Chem. 1993;268:6610-6614. 23. Seibert K, Zhang Y, Leahy K, et al. Pharmacological and biochemical demonstration of the role of cyclooxygenase 2 in inflammation and pain. Proc Nut1 Acad Sci IJSA. 1994;91:12013-12017. 24. Masferrer JL, Zweifel BS, Manning PT, et al. Selective inhibition of inducible cyclooxygenase 2 in vivo is anti-inflammatory and non-ulcerogenic. Proc Nut1 Acud Sci USA. 1994;9 1:3228-3232.

25. Chan CC, Boyce S, Brideau C, et al. Pharmacology of a selective cyclooxygenase2 inhibitor, L745.337-a novel nonsteroidal anti-inflammatory agent with an ulcerogenic sparing effect in rat and nonhuman primate stomach. J Phannacol Exp Ther. 1995;273:1531-1537. 26. Langman MJS, Weil J, Wainwright P, et al. Risks of bleeding peptic ulcer associated with individual non-steroidal anti-inflammatory drugs. Lancet. 1994;343: 1075-1078. 27. Fries JF. Assessing and understanding patient risk. Stand J Rheumatol Suppl. 1992;92:2 l-24. 28. Gabriel SE, Jaakkimainen L, Bombardier C. Risk for serious gastrointestinal complications related to use of nonsteroidal anti-inflammatory drugs. A meta-analysis. Ann Intern Med. 1991;15:787-796. 29. Hawkey CJ. COX-2 inhibitors. 1999;353:307-314.

Luncet.

30. Ehrich E, Dallob A, Van Hecken A, et al. Demonstration of selective COX-2 inhibition by MK-966 in humans. Arthritis Rheum. 1996:39(Suppl):S81. Abstract. 31. Depre M, Ehrich C, De Lepeleire I, et al. Demonstration of COX-2 inhibition by MK966 (Vioxx) in humans with supratherapeutic doses.Rheum EM 1998;27:196.Abstract. 32. Saag K, Fisher C, McKay J, et al. MK0966, a specific COX-2 inhibitor, has clinical efficacy comparable to ibuprofen in the treatment of knee and hip osteoarthritis (OA) in a 6 week controlled clinical trial. Arthritis Rheum. 1998;41 (Suppl):A984. Abstract. 33. Cannon G, Caldwell J, Holt P, et al. MK0966, a specific COX-2 inhibitor, has clinical efficacy comparable to diclofenac in the treatment of knee and hip osteoarthritis (OA) in a 26 week controlled clinical trial. Arthritis Rheum. 1998;41(Suppl):A983. Abstract.

1701

CLINICAL

34. Ehrich EW, Dallob AM, De Lepeleire I, et al. Characterization of rofecoxib as a cyclooxygenase-2 isoform inhibitor and demonstration of analgesia in the dental pain model. Clin Pharmacol Ther 1999; 65:336-347. 35. Morrison SW, Daniels SE, Kotey P et al. Rofecoxib, a specific cyclooxygenase-2 inhibitor, in primary dysmenorrhea: A randomized controlled trial. Obstet Gynecol. In press. 36. 1983 Metropolitan Height and Weight Tables. Stat Bull. 1983;64:3-9.

THERAPEUTICS”

drugs on the disposition of methotrexate in patients with rheumatoid arthritis. Biopharm Drug Dispos. 1998;19:163-167. 44. Davies NM, Anderson KE. Clinical pharmacokinetics of diclofenac. Clin Pharmacokinet. 1997;33:184-213. 45. Tracy TS, Worster T, Bradley JD, et al. Methotrexate disposition following concomitant administration of ketoprofen, piroxicam and flurbiprofen in patients with rheumatoid arthritis. Br J Clin Pharmacol. 1994:37:453-456.

37. Amett FC, Edworthy SM, Bloch DA, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum. 1996;39:723-731.

46. Combe B, Edno L, Lafforgue P, et al. Total and free methotrexate pharmacokinetits with and without piroxicam in rheumatoid arthritis patients. Br .I Rheumutol. 1995:34;421428.

38. Felson DT, Anderson JJ, Boers M, et al. American College of Rheumatology preliminary definition of improvement in rheumatoid arthritis. Arthritis Rheum. 199.5;38:727-735.

47. Wallace CA, Bleyer A, Sherry DD, et al. Toxicity and serum levels of methotrexate in children with juvenile rheumatoid arthritis. Arthritis Rheum. 1989;32:677-681.

39. Felson DT, Anderson JJ, Boers M, et al. The American College of Rheumatology preliminary core set of disease activity measures for rheumatoid arthritis clinical trials. Arthritis Rheum. 1993;36:729-740. 40. Barr DR. Using confidence intervals to test hypotheses. J Quul Technol. 1969; 1:256258. 41. Stein CM, Pincus T. Placebo-controlled studies in rheumatoid arthritis: Ethical issues. Lancet. 1999;353:40&403. 42. Rochon PA, Binns MA, Litner JA, et al. Are randomized control trial outcomes influenced by the inclusion of a placebo group? A systematic review of nonsteroidal antiinflammatory drug trials for arthritis treatment. J Clin Epidemiol. 1999; 52:113-122. 43. lqbal MP, Baig JA, Ali AA, et al. The effects of non-steroidal anti-inflammatory 1702

48. ACR Ad Hoc Committee on Clinical Guidelines. Guidelines for monitoring drug therapy in rheumatoid arthritis. Arthritis Rheum. 1996;39:723-73 1. 49. Bjarnason I, Sigthorsson G, Crane R, et al. COX-2 specific inhibition with MK0966 25 or 50 mg qd does not increase intestinal permeability: A controlled study with placebo (PBO) and indomethacin 50 mg tid (INDO). Am J Gastroenterol. 1998;93:A246. Abstract. 50. Hunt R, Bowen B, James C, et al. COX-2 specific inhibition with MK-0966 2.5mg or 50 mg qd over 4 weeks does not increase fecal blood loss: A controlled study with placebo and ibuprofen 800 mg tid. Am J Gastroenterol. 1998;93:A247. Abstract. 51. Whelton A, Hamilton CW. Nonsteroidal antiinflammatory drugs: Effects on kidney function. J Clin Pharmacol. 1991;31:588598.