Methods

Three large OA clinical trials meeting the following criteria were included in the integrated analysis: phase III development, parallel, randomized, placebo-controlled, double-blind, and multicenter studies conducted to evaluate the efficacy and safety of naproxcinod in patients with OA of the knee (2 studies) or hip (1 study). To be included in these 3 studies, the patients were required to be men or women ≥40 years old with a confirmed diagnosis of OA of the knee or hip in accordance with the American College of Rheumatology guidelines. They also had to be current chronic users of NSAIDs or acetaminophen for OA pain and to have experienced a flare of pain after a discontinuation period of ≥5 half lives of the previous analgesic or anti-inflammatory therapy. A flare was defined as a visual analog pain scale score at baseline of ≥50 mm and an increase of ≥15 mm compared to the screening visit. The patients were excluded if they had uncontrolled diabetes, clinically important liver or renal impairment (serum creatinine >176 μmol/L at screening), or gastrointestinal bleeding or ulceration within the previous 6 months. Patients with hypertension were included if their hypertension control at study entry was acceptable according to the judgment of the study physician and their antihypertensive drug regimen had been stable for ≥3 months.

In 2 of the 3 studies involving 1,938 patients with OA of the knee at 238 clinical sites in the United States, the patients were randomized in a 1:1:1:1 ratio to receive 1 of 4 treatments, comprising naproxcinod 750 mg, naproxcinod 375 mg, naproxen 500 mg, or placebo twice daily, for the 13-week study treatment period. In the third study, 810 patients with hip OA at 105 clinical sites in Europe and North America were randomized to receive naproxcinod 750 mg, placebo, or naproxen 500 mg twice daily in a 2:2:1 ratio. An unbalanced distribution of patients among the treatment groups occurred because naproxcinod 375 mg was not included in the hip OA study and a lesser number of patients were randomized to naproxen.

The focus of the present study was the effect of therapy on BP. The complete efficacy findings and other secondary end points for each of the 3 trials have been previously reported. To determine that no clinically relevant difference in pain control was observed among the active treatment groups, one of the co-primary end points for pain (Western Ontario and McMaster Universities Osteoarthritis Index visual analog scales for pain) is provided in the present report.

In each of the 3 studies, the seated clinic BP was measured in a standardized manner using mercury column or aneroid manometry in triplicate in the same arm and by the same examiner at baseline and 2, 6, and 13 weeks of the double-blind treatment period. The clinic BP was measured in the morning 2 to 4 hours after a dose of the study drug and, if applicable, antihypertensive drugs that had been taken in the morning. For patients treated with antihypertensive drugs, the interval between the intake of antihypertensive medication, the intake of the study medication, and the BP measurements remained the same throughout for each subject in all 3 studies. The times of medication dosing and BP measurements were recorded in the case report forms.

The comparability of the 4 treatment groups was determined from the demographic data and baseline BP values. Continuous variables were analyzed using the analysis of variance with treatment as the factor. Discrete variables were analyzed using the chi-square test.

The mean changes from baseline at week 13 in systolic and diastolic BP were assessed in a meta-analysis of the individual patient data using mixed linear models, with the study and treatment as categorical fixed effects and the baseline BP as a continuous fixed effect and assuming heterogeneous within-study variance. The 3 active treatment groups were compared to the placebo group using one-sided 95% confidence intervals (CIs) of the differences in the mean changes. The naproxcinod doses were compared to the naproxen dose using 2-sided 95% CIs. Post hoc, 2-sided CIs of the comparisons with placebo and the associated p values for all pairwise comparisons were calculated. Treatment-by-study interaction was added to the model as an additional random effect to test for heterogeneity across studies.

These analyses of the changes from baseline in BP during the study were performed for the entire safety population and for those patients taking renin-angiotensin system (RAS) inhibitors at baseline, including angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, β blockers, and renin inhibitors. The rationale for studying the RAS blocker subgroup was based on previous work, which showed a larger effect of NSAIDs on BP destabilization in patients taking these classes of antihypertensive agents. For the interpretation of the comparisons between the active treatment groups and the placebo group, a threshold of 2 mm Hg for the upper limit of the CI for the systolic BP differences and 1.5 mm Hg for the upper limit of the CI for the diastolic BP differences were used based on the relations to both cardiovascular outcomes and the clinical effect of these BP differences.

The distributions of the categorized systolic BP changes from baseline (increases in varying individual thresholds of BP) were calculated. The post hoc analyses of these categorized changes were done using the same fixed effect meta-analysis as described and performed on the log odds ratios estimated for each study individually in a logistic regression analysis. Treatment differences are presented using odds ratios, 2-sided 95% CIs, and associated p values.

The analyses were conducted post hoc to determine the estimated probability of destabilizing systolic BP (i.e., reaching a systolic BP of ≥140 mm Hg at week 13 given a baseline systolic BP of <140 mm Hg). In these analyses, the p values for the treatment comparisons were obtained from the pairwise contrasts from a logistic regression analysis with study and treatment as categorical fixed effects and baseline systolic BP as a continuous fixed effect. The mean constant curves were provided for graphic purpose and were determined from the results from the logistic regression analysis for the 3 individual studies.

Results

A total of 2,734 patients were included in the integrated safety population. As shown in Figure 1 , 2,079 of the patients (76.0%) completed the trial, with similar proportions in the active treatment groups. Fewer patients completed the study in the placebo group (70.8%), primarily owing to a lack of efficacy and the worsening of OA symptoms. The baseline characteristics of the study population are listed in Table 1 . Most patients were women and white, and the mean age of the study population was 61 years. The mean body mass index for the population was 32 kg/m ² and was similar across the treatment groups. Statistically significant differences were observed among the 4 treatment groups for race (black vs nonblack, p = 0.007), geographic location (Europe vs North America, p <0.0001), and body mass index (p <0.0001). A total of 1200 patients (44%) had both a history of hypertension and were receiving antihypertensive drug therapy. Of the patients treated with antihypertensive agents, 332 (27.7%) were taking a RAS inhibitor drug as a single agent for hypertension.

Entry, randomization, and follow-up of patients after randomization in integrated safety population.

In each of the 3 studies, all active treatments produced significant improvements from baseline in the Western Ontario and McMaster Universities Osteoarthritis Index pain subscale score at 13 weeks ( Table 2 ). The changes from baseline in the Western Ontario and McMaster Universities Osteoarthritis Index pain subscale score at week 13 were similar for naproxcinod 750 mg and naproxen 500 mg. Smaller reductions in the pain subscale score were seen for naproxcinod 375 mg versus naproxcinod 750 mg and naproxen 500 mg twice daily. Both doses of naproxcinod and naproxen 500 mg were significantly superior statistically to placebo ( Table 2 ). The number of patients who discontinued because of a lack of arthritis efficacy was similar for the active treatment groups (naproxcinod 750 mg, 4.9%; naproxcinod 375 mg, 6.7%, and naproxen 500 mg, 4.8%; Figure 1 ).

The changes from baseline in the systolic BP at each study visit during the 13-week period are shown in Figure 2 . The systolic BP in the naproxcinod 750-mg group decreased from baseline to a nadir of −3 mm Hg at week 6 and was of a similar magnitude to the BP changes in the placebo treatment group. In contrast, the systolic BP in the naproxen 500-mg group was unchanged from baseline during the 13-week period. Similar patterns of changes were seen for diastolic BP ( Figure 2 ). Additionally, differences in the mean changes from baseline to week 13 for both naproxcinod doses were similar to those with placebo (upper bound of 95% CI was 0.8 mm Hg and 2.1 mm Hg for naproxcinod 750 mg and 375 mg, respectively). The changes in systolic BP with naproxen 500 mg were significantly greater than those with placebo at week 13 (1.4 mm Hg, 95% CI 0.1 to 2.7, p = 0.032; Table 3 ). At weeks 6 and 13, the mean changes from baseline in diastolic BP were significantly lower for both doses of naproxcinod versus naproxen 500 mg twice daily (p <0.05 at both points).

Least square mean changes from baseline in (A) systolic BP (mm Hg) and (B) diastolic BP (mm Hg) during 13-week treatment period in integrated safety population. Number of patients enrolled shown at various points in trials.

Table 3

Least square mean change from baseline versus placebo and naproxen in systolic blood pressure

Integrated Safety Population	Naproxcinod 750 mg Twice Daily	Naproxcinod 375 mg Twice Daily	Naproxen 500 mg Twice Daily
Week 2
Comparison vs placebo
Patients (n)	783 vs 786	474 vs 786	623 vs 786
Mean difference (mm Hg)	0.18 (0.86 to 1.23)	0.72 (−0.53 to 1.97)	2.11 (0.99 to 3.23)
p Value	0.7302	0.2564	0.0002
Comparison vs naproxen
Patients (n)	783 vs 623	474 vs 623	—
Mean difference (mm Hg)	−1.93 (−3.04 to −0.81)	−1.39 (−2.66 to 0.12)	—
p Value	0.0007	0.0323	—
Week 6
Comparison vs placebo
Patients (n)	724 vs 693	445 vs 693	587 vs 693
Mean difference (mm Hg)	−0.13 (−1.24 to 0.96)	0.81 (−0.57 to 2.18)	2.60 (1.41 to 3.79)
p Value	0.8132	0.2486	<0.0001
Comparison vs naproxen
Patients (n)	724 vs 587	445 vs 587	—
Mean difference (mm Hg)	−2.73 (−3.91 to −1.55)	−1.79 (−3.17 to −0.41)	—
p Value	<0.0001	0.0110	—
Week 13
Comparison vs placebo
Patients (n)	662 vs 607	405 vs 607	522 vs 607
Mean difference (mm Hg)	−0.37 (−1.57 to 0.84)	0.67 (−0.79 to 2.13)	1.42 (0.12 to 2.71)
p Value	0.5505	0.3654	0.0321
Comparison vs naproxen
Patients (n)	662 vs 522	405 vs 522	—
Mean difference (mm Hg)	−1.78 (−3.05 to −0.51)	−0.74 (−2.20 to 0.72)	—
p Value	0.0059	0.3176	—

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Relation of Mitral Annular Calcium and Coronary Calcium (from the Multi-Ethnic Study of Atherosclerosis [MESA])
2. Coronary Flow: The Holy Grail of Echocardiography?
3. Real-Time Three-Dimensional Transesophageal Echocardiography for Assessment of Mitral Valve Functional Anatomy in Patients With Prolapse-Related Regurgitation
4. Relation Between Migraine and Size of Echocardiographic Intrapulmonary Right-to-Left Shunt

Stay updated, free articles. Join our Telegram channel

Join