Arcoxia for Arthritis Provides Limited GI Protection

LOS ANGELES, Feb. 9 -- Arthritis patients taking Arcoxia (etoricoxib), an investigational highly selective Cox-2 inhibitor, had significantly fewer uncomplicated upper gastrointestinal events, versus a traditional NSAID, but it was no better for more serious GI problems.

Moreover, the overall benefit for Arcoxia appears to have been small, according to results of a study in the Feb. 10 issue of The Lancet that pooled data from more than 35,000 arthritis patients taking Arcoxia or diclofenac.

The study's aim was to assess the effect of these drugs in a "real-world" situation among patients taking gastrointestinal protective therapy with proton pump inhibitors, said Loren Laine, M.D., of the University of Southern California here, and colleagues.

The gastrointestinal findings were a secondary outcome of the Multinational Etoricoxib and Diclofenac Arthritis Long-term (MEDAL) program, a pooled analysis of three double-blind randomized trials designed to assess the risks of thrombotic cardiovascular events among patients taking these drugs. These results were published in The Lancet in November http://www.medpagetoday.com/Rheumatology/Arthritis/tb/4500.

Uncomplicated upper GI events included uncomplicated bleeding or small ulcers generally not requiring hospitalization. Serious complicated events included perforation, obstruction, witnessed ulcer bleeding, or significant active bleeding, according to the investigators.

The Medal program included 34,701 patients with osteoarthritis (72%) or rheumatoid arthritis (28%). To simulate real-world practice, one-third of the patients took proton-pump inhibitors for gastrointestinal protection and one third took low-dose aspirin for cardiovascular protection.

Overall, upper gastrointestinal clinical events were significantly less common with Arcoxia (60 to 90 mg daily) than with diclofenac (150 mg daily), with a hazard ratio (HR) of 0.69 (95% CI 0.57-0.83; P=0.0001).

This benefit derived mainly from a reduction in uncomplicated events. There were significantly fewer uncomplicated gastrointestinal events with Arcoxia than there were with diclofenac (HR 0.57, 0.45-0.74; P

This decrease was similar in patients who took proton-pump inhibitors, suggesting that the COX-2 selective inhibitor provides symptomatic benefit even in patients already taking a PPI, the researchers said.

The rate of upper gastrointestinal uncomplicated clinical events in the MEDAL program was lower by about 60% than rates in other long-term outcome studies, the researchers said.

Nevertheless, neither these lower rates nor the concomitant use of proton-pump inhibitors and low-dose aspirin seem likely to fully explain the absence of a significant difference in serious upper gastrointestinal events, the researchers said.

In patients taking proton-pump inhibitors and low-dose aspirin, Arcoxia reduced the risk of uncomplicated upper gastrointestinal clinical events and dyspepsia compared with the traditional NSAID diclofenac, the researchers said. However, these risk reductions were seen only in the more common, but less serious, uncomplicated events, Dr. Laine's team wrote.

Summing up, the researches said, the results of the MEDAL program provide new information about upper gastrointestinal clinical events and symptoms to assist arthritis patients and their physicians in making decisions regarding NSAID use.

In an accompanying editorial, Joost Drenth, M.D., and Freek Verheug, M.D., of Radboud University in Nijmegen in The Netherlands, expressed skepticism about the significance of the results and concern about the role of "industry-driven" research in this study.

They pointed out the difficulty of drawing conclusions from a study designed to study a separate question. "The primary endpoint of the MEDAL program was thrombotic cardiovascular events," they wrote. "Although upper gastrointestinal toxicity was recorded, the program was not designed to address the question of whether etoricoxib limits gastrointestinal toxicity."

"To attribute benefits to secondary outcomes in a trial can be problematic. The selection of patients for the secondary outcome measure results in a different distribution over the groups than the selection for the primary endpoint," they added. "In other words, the groups of patients in the MEDAL program were allocated according to cardiovascular risk factors rather than upper gastrointestinal risk factors, introducing a possible source of bias."

The editorialists also questioned why the number of patients with significant complications did not differ in the two treatment groups. A drug that prevents uncomplicated disease, they said, might be expected to reduce serious side effects such as bleeding and perforation. However, they wrote, it did not, and this paradox has been found in other Cox-2 studies. Yet because peptic ulcer complications are relatively rare, the MEDAL program may have been underpowered to detect an actual treatment effect.

To suggest what the true treatment effect might be, Drs. Drenth and Verheugt calculated that it would be necessary to treat 259 patients with Arcoxia to prevent one uncomplicated gastrointestinal event in one patient. So although the effect might be statistically significant, they said, the effect is not large and might not be clinically relevant.

Addressing the potentially heavy hand of industry in the MEDAL program, they wrote that the statistical analysis, key to the results, was done by employees from Merck Research Laboratories, maker of Arcoxia. The company had influence over all aspects, including data analysis, safety monitoring, and reporting. This control might not necessarily affect the credibility of the results, they said, but the demise of rofecoxib (Vioxx) indicates that independent data analysis is desirable.

The real question, the editorialists wrote, is whether a COX-2 inhibitor, such as Arcoxia, is safer than a proton-pump inhibitor added to a standard non-steroidal anti-inflammatory drug. The advantage of the proton-pump inhibitor option is that it is less expensive, potentially less cardiotoxic, and advantageous in terms of reducing dyspepsia, but confirmation of this needs a new randomized trial.

A Cochrane review supports the safety of add-on proton-pump inhibitor, and an observational study suggests that this strategy has a similar safety profile to Cox-2 inhibition alone, they said. The addition of the proton-pump inhibitor in this study may have led to a non-random selection of groups, leading to an erroneous conclusion. However, the nature of the MEDAL program does not allow a definitive answer to that question. "As such, Laine and colleagues' article merely generates the hypothesis rather than answering it," they concluded.

Dr. Laine reported receiving research support from Merck, Pfizer, Novartis, TAP, and Bayer, and has served as a consultant for Merck, Novartis, Bayer, AstraZeneca, Eisai, Altana, J&J, and Santarus. Sean Curtis, M.D., and Amarjoy Kaur, M.D., are employees of Merck and own stock and/or hold stock options in the company. Byron Cryer, M.D., has served as a consultant for Merck, Pfizer, AstraZeneca, and TAP. Christopher Cannon, M.D., has received research grant support from Accumetrics, AstraZeneca, Merck, Merck/Schering Plough Partnership, and Schering Plough through the Department of Medicine of Brigham and Women's Hospital and has served on scientific advisory boards for Alnylam, AstraZeneca, Biosite, Bristol-Myers Squibb, Eisai, GlaxoSmithKline, Merck, Merck Schering Plough Partnership, Pfizer, Sanofi-Aventis, Schering Plough, and Tethys. Dr. Cannon also reported receiving lecture fees from Accumetrics, AstraZeneca, Bristol-Myers Squibb, Merck, Merck/Schering Plough Partnership, Pfizer, Sanofi-Aventis, and Schering Plough; he has received honoraria for the preparation of educational materials from BGB New York, DIME, and NCME; and has received additional funding for studies and subsidies conducted by the TIMI study group from Merck, Bristol-Myers Squibb, Sanofi-Aventis, Millennium Pharmaceuticals, Nuvelo, AstraZeneca Pharmaceuticals, CV Therapeutics, Inotek Pharmaceuticals, Bayer Healthcare LLC, Ortho-Clinical Diagnostics, Sanofi-Synthelabo Recherche, GlaxoSmithKline, Amgen, Beckman Coulter, Biosite Coulter, Biosite Incorporated, Roche Diagnostics, Roche Diagnostics GmbH, Pfizer, Accumetrics, and Novartis Pharmaceuticals.