COX-2 Inhibitor - Comparative Studies

Comparative Studies

In a metaanalysis of eight osteoarthritis studies, the incidence of withdrawal because of adverse gastrointestinal events was 3.5% for Vioxx, compared to 4.8% for ibuprofen, diclofenac, or nabumetone (Relafen). Endoscopic studies of patients receiving Celebrex 50–400 mg twice daily for 12–24 weeks found rates of upper gastrointestinal complications similar to placebo and significantly lower than naproxen 500 mg twice daily and ibuprofen 800 mg three times daily, but not statistically significantly different from patients receiving diclofenac 75 mg twice daily.11 The analysis found that Vioxx provided significant gastrointestinal benefits in patients both at high risk and at low risk of developing gastrointestinal problems; patients at low risk still had 88% fewer gastrointestinal problems with Vioxx.

The results of the CLASS study were confirmed by the Successive Celecoxib Efficacy and Safety Studies (SUCCESS) study, which examined the effectiveness and safety of celecoxib 200 mg and 400 mg daily and how well it was tolerated by patients in terms of adverse effects, compared with the most common NSAID regimens in the countries studied (diclofenac 100 mg daily and naproxen 1000 mg daily). SUCCESS showed that celecoxib was as effective as the conventional NSAIDs in controlling the pain of arthritis, and caused fewer gastrointestinal ulcers or ulcer complications (such as perforations or bleeding) and fewer upper gastrointestinal adverse effects, e.g. 29% less chance of having nausea and 22% less chance of abdominal pain. In addition, hospitalization rates for upper gastrointestinal problems were 2 to 4 times lower with celecoxib, and because there were fewer adverse effects, there was 23% less chance of a celecoxib patient stopping treatment. The study also found that there was no real advantage to taking a bigger dose of celecoxib: the 200 mg dose was found to be just as effective as the 400 mg dose.

The VIGOR study was followed by the Assessment of Difference between Vioxx and Naproxen to Ascertain Gastrointestinal Tolerability and Effectiveness (ADVANTAGE) study, which showed that 9.1% of people taking Vioxx received a gastro-protective medicine compared with 11.2% of people taking naproxen, a reduction of 19%. In addition, after 3 months, 5.9% of people stopped taking Vioxx compared with 8.1% who stopped taking naproxen, a reduction of 27%. ADVANTAGE was the first study comparing the gastrointestinal tolerability of Vioxx and naproxen in a group that included patients taking low-dose aspirin for cardiovascular reasons. This was followed by the Experience with Vioxx in Arthritis (EVA) survey of 5,986 Belgian physicians and 74,192 people with osteoarthritis, which found that, after 12.5 or 25 mg of Vioxx once daily for 30 days, 80% of the patients wished to continue treatment with Vioxx and more than 80% of doctors said they would continue prescribing Vioxx. The preference to continue taking Vioxx was especially strong in people previously treated with older NSAIDs.

In a six week long study comparing Vioxx 25 mg once daily, Vioxx 12.5 mg once daily, Celebrex 200 mg once daily, and paracetamol/acetaminophen 1,000 mg four times daily for osteoarthritis of the knee, higher dose Vioxx was found to be superior to the other three treatments for reduction of nocturnal pain, and superior to Celebrex and acetaminophen/paracetamol for reduction of resting pain. At six weeks, 60% of high dose Vioxx patients reported a good or excellent response, compared to 46% of Celebrex patients and 39% of paracetamol patients. Low dose Vioxx was not found to be statistically significant from Celebrex at this dose. Similar results were found for early response to therapy.

However, when the Food and Drug Administration (FDA) later presented more complete data from the CLASS and VIGOR trials on its web site, the results were less certain. The CLASS trial was revealed to also have twelve and fifteen month time points which had not been discussed in the JAMA publication; in this segment of the trial, the number of ulcer-related complications for Celebrex caught up to the control NSAID group. Similarly, the complete VIGOR study data revealed that in fact, when all adverse events, not just gastrointestinal, were tabulated, the patients receiving VIOXX had suffered (barely) significantly higher incidence of adverse events overall than the control NSAID group. In particular, the risk of serious cardiovascular thrombotic events, e.g. myocardial infarction, was 1.7% in the VIOXX patients versus 0.7% in the control group, and there were significantly more withdrawals in the Vioxx group for causes including hypertension, edema, hepatotoxicity, heart failure, or pathological laboratory findings. The mean increases in systolic and diastolic blood pressure in the Vioxx group were 4.6 mmHg and 1.7 mmHg respectively, compared to 1.0 and 0.1 mmHg in the control NSAID group. An estimated 43,000,000 Americans, nearly one out of six, suffers from arthritis. However, 42% (18 million) of these also suffer from hypertension. Therefore, the promise of better patient outcomes and lowered medical costs from use of COX-2 inhibitors may not be as great as previously hoped. Questions remain regarding the relative safety and cost effectiveness of this new class. While endoscopic evidence of gastrointestinal damage is frequently seen in studies of nonspecific NSAIDs, the actual incidence of clinically evident symptoms and patient discomfort is much lower; furthermore, in cases of short-term therapy, any such damage generally reverses itself quickly after termination of the drug.