Evidence in practice — number 3: Cox 2 inhibitors

Clinical question — Are second generation Cox-2 inhibitors, such as lumiracoxib, a sensible and safe alternative to naproxen or ibuprofen in patients with osteoarthritis?

The evidence. Schnitzer TJ, Burmester GR, Mysler E, et al. Comparison of lumiracoxib with naproxen and ibuprofen in the Therapeutic Arthritis Research and Gastro-intestinal Event Trial (TARGET), reduction in ulcer complications: randomised controlled trial. Lancet 2004; 364(9435): 665–674.1

Farkouh ME, Kirshner H, Harrington RA, et al. Comparison of lumiracoxib with naproxen and ibuprofen in the Therapeutic Arthritis Research and Gastrointestinal Event Trial (TARGET), cardiovascular outcomes: randomised controlled trial. Lancet 2004; 364(9435): 675–684.2

Background. In 2001 the National Institute of Clinical Excellence produced guidance on the use of the first generation Cox-2 inhibitors (celecoxib, rofecoxib, meloxicam, and etodolac) and suggested that these drugs could have a role in patients requiring a non-steroidal anti-inflammatory drug (NSAID) who may be at ‘high risk’ of developing gastrointestinal adverse effects. However, some concerns were expressed about the prescription of Cox-2s in patients with cardiovascular disease, particularly those on low dose aspirin. In the VIGOR trial refecoxib was associated with an increase in myocardial infarctions compared with naproxen.3

Study design and intervention. TARGET was a double-blind, double-dummy parallel group randomised, controlled trial. In such a design, each group of participants receives one of the active interventions and a placebo (in this case called a dummy) that looks and tastes the same as the alternative intervention. The study involved 18 325 patients across 29 countries and was 90% powered to detect a 50% reduction in definite or probable upper gastrointestinal ulcer complications in patients not taking low-dose aspirin. The study was divided into two similar sized sub-studies groups, which were stratified, before randomisation, by age and aspirin use. One sub-study compared lumiracoxib 400mg daily with ibuprofen 800mg three times daily; the second compared lumiracoxib 400mg daily with naproxen 500mg twice daily. To be included, patients had to be aged 50 years or older and have a clinical diagnosis of osteoarthritis of the hip, knee, or hand, or radiographic evidence of cervical or lumbar spine osteoarthritis. They also had to be in moderate pain and likely to require treatment for at least a year.

Patients were excluded if they were taking gastroprotective drugs and/or had a history of gastointestinal ulceration (in the last 30 days), upper gastrointestinal bleeding (in the last year), or gastroduodenal perforation or obstruction. Patients were also excluded if they had a history of significant cardio-vascular disease encompassing myocardial infarction, stroke, coronary surgical inter-ventions, or new-onset angina (within the previous 6 months). However, 46% of the study population had hypertension, 20% dyslipidaemia, 8% diabetes, and 10% were current smokers. Of patients, 13% were classified as being at elevated cardiovascular risk (2% from Framingham risk scoring and 13% by virtue of a history of vascular disease). In terms of cotreatments, 24% were on aspirin, 12% on β-blockers and 26% on drugs affecting the renin-angiotensin system.

Outcomes and analysis. The primary endpoint was the cumulative 1-year incidence of definite or probable upper gastrointestinal ulcer complications (defined as clinically significant bleeding, perforation, or obstruction from erosive or ulcer disease).

In relation to cardiovascular outcomes, the investigators sought to identify all patients with confirmed or probable cardiovascular death, myocardial infarction, or stroke. Together these outcomes constituted the ‘composite cardiovascular endpoint’.

Results. Significantly fewer patients on lumiracoxib (29 [0.32%]) developed upper gastrointestinal complications compared with 83 (0.91%) on ibuprofen/naproxen. The difference was also significantly in favour of lumiracoxib for the individual comparisons between drugs (that is, naproxen versus lumiracoxib P = 0.0002 and ibuprofen versus lumiracoxib P = 0.0006). These significant differences were maintained in the sub-group of patients not taking aspirin, but not for patients also taking aspirin.

No significant differences were recorded in the overall composite cardiovascular endpoint for lumiracoxib versus ibuprofen/naproxen. Focusing down on myocardial infarction it seems that, compared with lumiracoxib, fewer patients on naproxen sustained an infarction and more patients on ibuprofen infarcted. However, neither of these differences achieved statistical significance.

Commentary. Considering the TARGET trial in context, it is important to be aware that all NSAIDs (including Cox-2s) have side-effects and the decision on which variety to select should be reserved until after a decision to prescribe an NSAID-type treatment has been made. Clearly, alternative management approaches will also require consideration, such as analgesics, physical treatments, and referral. From a population-based perspective, it is self-evident that over-enthusiastic prescription of an NSAID (including a Cox-2) with a lower level of gastrointestinal side effects could easily result in an increase in the overall numbers of adverse gastrointestinal events (and with more costly prescribing) in comparison with the more restricted prescribing of an NSAID with a higher level of gastrointestinal adverse effects.

Over the course of a year, patients may discontinue prescribed NSAIDs for a variety of reasons — this must be borne in mind when reviewing any trial of NSAID-type treatment. In TARGET, over a third of patients failed to complete the trial with significantly more of these ‘lost’ patients (248) being those treated with ibuprofen/ naproxen than with lumiracoxib. As the analysis focused on all patients randomised who received at least one dose of study drug, this places lumiracoxib at a slight statistical disadvantage compared with ibuprofen/ naproxen … adding some further weight to the study findings.

The bottom line. In patients over the age of 50 years with osteoarthritis, lumiracoxib showed a three- to fourfold reduction in ulcer complications compared with naproxen/ ibuprofen without a significant increase in the rate of serious cardiovascular events.